1. Why TimveroOS?
Building financial applications traditionally means months of development for basic features like payment processing, KYC compliance, and risk management. TimveroOS changes this by providing production-ready financial components that you customize through code, not configuration.
1.1. The Problem
-
6-12 months to build basic lending platform from scratch
-
Complex integrations with payment gateways, KYC providers, credit bureaus
-
Regulatory compliance requirements that slow development
-
Scalability challenges as your business grows
1.2. The TimveroOS Solution
-
2-4 weeks to customize and deploy with pre-built components
-
Built-in integrations for payments, documents, notifications
-
Compliance frameworks for KYC, AML, reporting built-in
-
Enterprise-scale architecture from day one
1.3. What You’ll Build
This SDK empowers developers to create financial applications through: * Entity-Form-Controller Pattern - Define business objects with automatic CRUD operations * Comprehensive Form System - Handle complex forms with validation, nested components, and MapStruct mapping * Credit Operations Framework - Manage loan lifecycle with automated calculations, accruals, and payments * Payment Transaction System - Process real-world payments through multiple gateways with complete audit trails * Entity Checkers - Implement event-driven business rules that respond to data changes automatically
For a complete working example, see the timvero-example project.
2. Learning Path
This guide is organized into logical parts that build on each other:
2.1. PART I: FOUNDATION
-
Platform Overview - Core concepts: Basic, Origination, and Servicing layers
-
Getting Started - Master the Entity-Form-Controller pattern with Client example
-
Data model setup - SQL autogeneration and Flyway migration workflows
2.2. PART II: USER INTERFACE
-
Form classes setup and usage - Form classes, validation, MapStruct mappers, and service layers
-
Form Templates - Form structure, layout patterns, and two-column grids for modal forms
-
HTML Template Integration - Thymeleaf components, validation classes, and UI integration
-
Fast Search System - Unified search across all entities with Hibernate Search and Lucene
-
Document Management - Document upload, requirements, and UI integration
-
Document Templates - Generate contracts, reports, and formatted documents
-
Notifications - Multi-channel automated notifications with template-based content === PART III: BUSINESS LOGIC
-
[entity-checkers-setup-and-usage] - Event-driven business rules and automation
-
DataSource Integration - External API integration and data enrichment
-
Workflow Integration - Integration for decision workflows
-
Product Web Components Configuration - Automated web component setup for credit products and additives
-
Offer Engine & Credit Products - Generate personalized loan offers from credit products and participant data
-
REST API Integration - REST API setup and external system integration
-
Docusign Integration - Document signing integration with DocuSign === PART IV: FINANCIAL OPERATIONS
-
Credit Management System - Complete credit lifecycle management
-
Credit Operations Framework - Credit operations: charges, payments, accruals, past due processing
-
Payment Transactions Framework - Real-world payment processing with gateway integration
| New to TimveroOS? Start with Part I and follow sequentially. Experienced developers can jump to specific parts based on their immediate needs. |
3. Platform Overview
Build lending applications fast with pre-built components.
3.1. What You Get
-
Forms — Collect customer data with built-in validation
-
Database — Automatic setup for loan data
-
Payments — Connect to banks and card processors
-
UI Components — Works on desktop and mobile
-
Business Rules — Automate loan decisions
3.2. What You Build
-
🎯 Your loan products (personal loans, mortgages, etc.)
-
🎯 Your approval rules (credit scores, income requirements, etc.)
-
🎯 Your customer experience (application flow, portal design, etc.)
3.4. How It Works
Basic (foundation) → Origination (new loans) → Servicing (active loans)
Example: Personal loan process
-
Customer applies → forms, documents (Basic + Origination)
-
Loan approved → credit check, terms (Origination)
-
Loan created → account setup (Origination → Servicing)
-
Customer pays → balance updates (Servicing + Basic)
3.5. Key Platform Features
The platform provides several key features that make building lending applications easier:
1. Consistent Patterns
Everything follows the same pattern, so once you learn one part, you understand the whole platform:
-
Entities - Store your business data (customers, loans, payments)
-
Forms - Collect information from users with built-in validation
-
Controllers - Handle user actions like creating or editing records
-
Templates - Display information to users in web pages
Example: Whether you’re working with customers, loan applications, or payments, they all follow the same pattern.
2. Automatic Business Rules
The platform can automatically handle business logic for you:
-
When a loan application is approved → automatically create the loan account
-
When a payment is received → automatically update the loan balance
-
When a payment is late → automatically calculate late fees
-
When a loan is paid off → automatically close the account
Example: You define the rule "charge a $25 late fee if payment is 10 days late" and the platform handles it automatically.
3. Complete Audit Trail
Everything is tracked automatically for compliance and troubleshooting:
-
Who made each change and when
-
What the data looked like before and after changes
-
Why each change happened (payment received, fee charged, etc.)
-
Complete history of every loan account
Example: You can see that on March 15th, John Smith made a $500 payment, which reduced his loan balance from $5,000 to $4,500.
3.6. What You Need to Know
To get started with the platform, you need to understand a few key concepts:
1. Everything Builds on Basic
No matter what you’re building, you’ll always use Basic components:
-
Every application needs forms to collect data
-
Every application needs a database to store information
-
Every application needs templates to show information to users
Start here: Learn the Basic components first, then move to Origination or Servicing.
2. Choose Your Focus
Decide what type of application you’re building:
Building a loan application system? → Focus on Origination
-
Customer registration and applications
-
Credit scoring and risk assessment
-
Loan approval workflows
Building a loan management system? → Focus on Servicing
-
Payment processing and account management
-
Interest calculations and late fees
-
Customer account portals
Building both? → Start with Origination, then add Servicing
3. The Platform Does the Hard Work
You don’t need to build everything from scratch:
-
Database setup is automatic
-
Form validation is built-in
-
Payment processing is pre-built
-
Interest calculations are handled for you
-
Audit trails are automatic
Focus on: Your business rules, your loan products, and your customer experience.
3.7. Next Steps
Now that you understand the platform basics:
-
Start with Getting Started - Set up your first application and see how it works
-
Learn the Basic Components - Master forms, database, and templates (you’ll use these everywhere)
-
Pick your focus:
-
For loan applications: Learn about customers, applications, and risk assessment
-
For loan management: Learn about credits, payments, and operations
-
-
Build your application - Start simple and add features as you learn
The platform handles the complex technical details so you can focus on your business logic and customer experience.
4. Getting Started
This section provides a quick introduction to building applications with the Timvero SDK. Follow this guide to get up and running in minutes and understand the core concepts through practical examples.
4.1. Prerequisites
Before you begin, ensure you have the following installed:
-
Java 21 or later - The platform requires modern Java features
-
Maven 3.8+ - For dependency management and building
-
PostgreSQL 16+ - Primary database for the platform
-
IDE with Spring Boot support - IntelliJ IDEA, Eclipse, or VS Code
4.2. Quick Setup (5 minutes)
Step 1: Clone and Configure
-
Clone the example project:
git clone https://github.com/TimveroOS/timvero-example.git cd timvero-example -
Configure database connection in
src/main/resources/application.properties:spring.datasource.url=jdbc:postgresql://localhost:5432/your_database spring.datasource.username=your_username spring.datasource.password=your_password -
Run the application:
mvn spring-boot:run -
Access the application:
-
Admin UI: http://localhost:8081
-
Portal API: http://localhost:8082
-
Step 2: Verify Installation
Once running, you should see:
-
Database tables automatically created via Flyway migrations
-
Admin interface with navigation menu
-
Sample data (if configured)
-
No compilation errors in the console
Troubleshooting Setup Issues
Application Won’t Start
Problem: Application failed to start with database connection errors
Solution:
-
Check database is running:
# PostgreSQL status check sudo systemctl status postgresql # Or for Docker docker ps | grep postgres -
Verify database connection:
# Test connection manually psql -h localhost -p 5432 -U your_username -d your_database -
Check application.properties:
# Ensure these match your database setup spring.datasource.url=jdbc:postgresql://localhost:5432/your_database spring.datasource.username=your_username spring.datasource.password=your_password
Common database URL mistakes:
-
Wrong port (default PostgreSQL is 5432)
-
Database name doesn’t exist
-
User lacks permissions
Flyway Migration Errors
Problem: FlywayException: Migration checksum mismatch or migration failures
Solution:
-
Check migration file integrity:
-- View migration history SELECT * FROM flyway_schema_history ORDER BY installed_rank DESC; -
Reset migrations (development only):
# WARNING: This deletes all data mvn flyway:clean flyway:migrate -
Skip problematic migration (careful!):
# Only if you understand the implications mvn flyway:repair
Port Already in Use
Problem: Port 8081 was already in use or similar port conflicts
Solution:
-
Find what’s using the port:
# Linux/Mac lsof -i :8081 # Windows netstat -ano | findstr :8081 -
Change application ports:
# In application.properties server.port=8090 management.server.port=8091 -
Kill conflicting process:
# Linux/Mac (replace PID with actual process ID) kill -9 PID # Windows taskkill /PID PID /F
Java Version Issues
Problem: UnsupportedClassVersionError or compilation failures
Solution:
-
Check Java version:
java -version javac -version echo $JAVA_HOME -
Ensure Java 21+:
# Install Java 21 if needed # Ubuntu/Debian sudo apt install openjdk-21-jdk # macOS with Homebrew brew install openjdk@21 -
Set JAVA_HOME:
# Linux/Mac - add to ~/.bashrc or ~/.zshrc export JAVA_HOME=/usr/lib/jvm/java-21-openjdk # Windows - set in System Properties
Maven Build Failures
Problem: Maven dependency resolution or compilation errors
Solution:
-
Clean and rebuild:
mvn clean compile mvn clean install -U # Force update dependencies -
Check Maven version:
mvn --version # Should be 3.8+ -
Clear local repository:
# Nuclear option - deletes all cached dependencies rm -rf ~/.m2/repository mvn clean install
Can’t Access Admin UI
Problem: Browser shows "This site can’t be reached" or connection refused
Solution:
-
Verify application started successfully:
# Check logs for "Started Application in X seconds" tail -f logs/application.log -
Check correct URL:
# Default URLs Admin UI: http://localhost:8081 Portal API: http://localhost:8082 # NOT http://localhost:8080 (that's often Spring Boot default) -
Check firewall/network:
# Test port connectivity telnet localhost 8081 # Or curl -I http://localhost:8081
Database Tables Not Created
Problem: Application starts but database is empty
Solution:
-
Verify migration files exist:
ls -la src/main/resources/db/migration/ # Should see V*.sql files -
Check database permissions:
-- User needs privileges to execute migration ALTER DATABASE your_database OWNER TO your_username;
4.3. Your First Entity: Client Management (15 minutes)
Let’s explore how the Client entity demonstrates the platform’s core patterns. The Client entity is already implemented in the example project, so you can see a complete working example.
Entity Definition
The Client entity demonstrates the platform’s entity structure:
@Entity
@Table
@Audited
@Indexed
public class Client extends AbstractAuditable<UUID> implements NamedEntity, HasDocuments {
@Embedded
@Valid
private IndividualInfo individualInfo;
@Embedded
@Valid
private ContactInfo contactInfo;
// getters and setters...
}
Key features:
-
Extends
AbstractAuditable: Automatic creation/modification tracking -
Implements
NamedEntity: Provides display name functionality -
Composite structure: Contains
IndividualInfoandContactInfocomponents -
Search integration:
@Indexedenables full-text search -
Audit support:
@Auditedtracks all changes
Form Structure
The ClientForm handles user input with validation:
public class ClientForm {
@Valid
@NotNull
private IndividualInfoForm individualInfo;
@Valid
@NotNull
private ContactInfoForm contactInfo;
// getters and setters...
}
Benefits:
-
Nested validation:
@Validcascades validation to nested objects -
Clean separation: Form objects separate from entities
-
Type safety: Strongly typed form fields
Controller Implementation
The main controller handles entity management:
@Controller
public class ClientController extends EntityController<UUID, Client, ClientForm> {
// Inherits all CRUD functionality automatically
}
Actions provide specific operations (buttons in the UI):
@Controller
public class CreateClientAction extends EntityCreateController<UUID, Client, ClientForm> {
@Override
protected boolean isOwnPage() {
return false;
}
}
@Controller
public class EditClientAction extends EditEntityActionController<UUID, Client, ClientForm> {
// Handles the edit button functionality
}
-
{ok} Create, Read, Update, Delete operations
-
{ok} Form validation and error handling
-
{ok} List view with search and filtering
-
{ok} Responsive web interface
-
{ok} Audit logging of all changes
Form Service Layer
The service layer handles business logic and data mapping:
@Service
public class ClientFormService extends EntityFormService<Client, ClientForm, UUID> {
// Inherits entity-form mapping and persistence operations
}
The service requires a corresponding MapStruct mapper for entity-form conversion:
@Mapper
public interface ClientFormMapper extends EntityToFormMapper<Client, ClientForm> {
// MapStruct automatically generates implementation for bidirectional mapping
}
Template Integration
The HTML template demonstrates the form component system:
<th:block th:insert="~{/form/components :: text(
#{client.individualInfo.fullName},
'individualInfo.fullName',
'v-required v-name')}" />
<th:block th:insert="~{/form/components :: text(
#{client.contactInfo.email},
'contactInfo.email',
'v-required v-email')}" />
4.4. Essential Concepts (10 minutes)
Entity-Form-Controller Pattern
The platform follows a consistent architectural pattern:
| Component | Purpose | Example |
|---|---|---|
Entity |
JPA entity with business logic and relationships |
|
Form |
DTO for user input with validation rules |
|
Controller |
Main entity controller providing CRUD operations |
|
Actions |
Specific operation buttons in the UI |
|
Service |
Business logic and entity-form mapping |
|
Mapper |
Automatic bidirectional object mapping |
|
Automatic Features
Once you create the basic structure following this pattern, the platform automatically provides:
-
CRUD Operations: Complete create, read, update, delete functionality
-
Form Validation: Client-side and server-side validation
-
Database Migrations: Automatic schema generation and versioning
-
Search and Filtering: Full-text search and advanced filtering
-
Audit Logging: Complete change history tracking
-
Responsive UI: Mobile-friendly web interface
-
Security Integration: Authentication and authorization
-
API Endpoints: RESTful API for external integration
Data Flow
Understanding the data flow helps you work effectively with the platform:
User Input → Form Validation → Controller → Service → Mapper → Entity → Database
↓
Template Rendering ← Form Object ← Mapper ← Entity ← Database Query
Troubleshooting Entity-Form-Controller Issues
Form Validation Not Working
Problem: Form submits with invalid data or validation messages don’t appear
Solution:
-
Check validation annotations:
// Ensure @Valid is present on nested objects @Valid @NotNull private IndividualInfoForm individualInfo; -
Verify form component validation classes:
<!-- Ensure validation CSS classes are included --> <th:block th:insert="~{/form/components :: text( #{client.individualInfo.fullName}, 'individualInfo.fullName', 'v-required v-name')}" />
Controller Actions Not Appearing
Problem: Create/Edit buttons don’t show up in the UI
Solution:
-
Check controller annotations:
@Controller // Must be @Controller, not @RestController public class CreateClientAction extends EntityCreateController<UUID, Client, ClientForm> { } -
Check template includes actions:
<!-- Ensure action templates are included --> <div th:replace="~{/entity/actions :: entityActions}"></div>
Form Fields Not Displaying
Problem: Form renders but specific fields are missing or empty
Solution:
-
Check form component syntax:
<!-- Ensure proper Thymeleaf fragment syntax --> <th:block th:insert="~{/form/components :: text( #{client.individualInfo.fullName}, 'individualInfo.fullName', 'v-required v-name')}" /> -
Verify i18n message keys exist:
# In messages.properties client.individualInfo.fullName=Full Name
4.5. Common Scenarios (20 minutes)
Adding Custom Validation
Enhance the Client form with custom business rules:
public class ClientForm {
@NotBlank
@Size(min = 2, max = 100, message = "Name must be between 2 and 100 characters")
private String fullName;
@NotBlank
@Email(message = "Please provide a valid email address")
private String email;
@NotBlank
@Phone(message = "Please provide a valid phone number")
private String phone;
@PastOrPresent(message = "Birth date cannot be in the future")
private LocalDate dateOfBirth;
}
Implementing Business Logic with Entity Checkers
Create automated workflows that respond to client changes:
@Component
public class ClientWelcomeChecker extends EntityChecker<Client> {
@Override
protected void registerListeners(CheckerListenerRegistry<Client> registry) {
// Trigger when a new client is created
registry.entityChange().inserted();
}
@Override
protected boolean isAvailable(Client client) {
// Only for clients with complete contact information
return client.getContactInfo() != null
&& client.getContactInfo().getEmail() != null;
}
@Override
protected void perform(Client client) {
// Send welcome email to new clients
emailService.sendWelcomeEmail(client);
log.info("Welcome email sent to client: {}", client.getIndividualInfo().getFullName());
}
}
Adding Document Management
Enable clients to upload required documents:
// 1. Make Client support documents
@Entity
public class Client extends AbstractAuditable<UUID> implements HasDocuments {
// Existing client implementation
}
// 2. Configure document types
@Configuration
public class ClientDocumentConfiguration {
public static final EntityDocumentType ID_DOCUMENT = new EntityDocumentType("ID_DOCUMENT");
public static final EntityDocumentType PROOF_OF_ADDRESS = new EntityDocumentType("PROOF_OF_ADDRESS");
@Bean
DocumentTypeAssociation<Client> clientRequiredDocuments() {
return DocumentTypeAssociation.forEntityClass(Client.class)
.required(ID_DOCUMENT)
.required(PROOF_OF_ADDRESS)
.build();
}
}
// 3. Add document management tab
@Controller
@Order(1000)
public class ClientDocumentsTab extends EntityDocumentTabController<Client> {
@Override
public boolean isVisible(Client client) {
return true; // Always show documents tab for clients
}
}
Integrating External Data Sources
Fetch additional data from external APIs:
// 1. Create a data source subject interface
public interface CreditCheckSubject {
String getNationalId();
String getFullName();
}
// 2. Implement the interface in your entity
@Entity
public class Client implements CreditCheckSubject {
@Override
public String getNationalId() {
return getIndividualInfo().getNationalId();
}
@Override
public String getFullName() {
return getIndividualInfo().getFullName();
}
}
// 3. Create the data source implementation
@Service("creditCheck")
public class CreditCheckDataSource implements MappedDataSource<CreditCheckSubject, CreditReport> {
@Override
public Class<CreditReport> getType() {
return CreditReport.class;
}
@Override
public Content getData(CreditCheckSubject subject) throws Exception {
// Call external credit check API
String response = creditCheckApi.checkCredit(
subject.getNationalId(),
subject.getFullName()
);
return new Content(response.getBytes(), MediaType.APPLICATION_JSON_VALUE);
}
@Override
public CreditReport parseRecord(Content data) throws Exception {
return objectMapper.readValue(data.getData(), CreditReport.class);
}
}
4.6. What’s Next?
Explore Advanced Features
Now that you understand the basics, dive deeper into specific areas:
-
Form Classes - Complex validation, nested forms, and custom components
-
Entity Checkers - Business rule automation and workflow triggers
-
Document Management - File uploads, document requirements, and digital signatures
-
DataSource Integration - External API integration and data enrichment
-
Template System - Custom UI components and advanced templating
Real-World Implementation Patterns
Study these complete examples in the project:
-
Client Onboarding: Complete customer registration with validation and document collection
-
Application Processing: Multi-step loan application workflow with automated decision making
-
Participant Management: Complex participant relationships with role-based permissions
-
Document Workflows: Digital signature processes with DocuSign integration
-
Risk Assessment: External data integration for credit scoring and fraud detection
Development Best Practices
-
Start Simple: Begin with basic CRUD operations, add complexity gradually
-
Follow Patterns: Use the established Entity-Form-Controller pattern consistently
-
Leverage Automation: Use Entity Checkers for business rules instead of manual processes
-
Test Thoroughly: The platform provides excellent testing support for all components
-
Monitor Performance: Built-in metrics and logging help optimize your application
Getting Help
-
Documentation: This guide covers all platform features in detail
-
Example Project: Every feature demonstrated with working code
-
Professional Support: Enterprise support available for production deployments
Next Steps Checklist
-
{todo} Create your first custom entity following the Client pattern
-
{todo} Add custom validation rules to your forms
-
{todo} Implement an Entity Checker for business logic automation
-
{todo} Set up document management for your entities
-
{todo} Integrate with an external data source
-
{todo} Customize the UI templates for your specific needs
-
{todo} Deploy to a staging environment for testing
You’re now ready to build powerful financial applications with the Timvero platform!
Next Chapter: Data model setup - SQL autogeneration and Flyway migration workflows
Related Chapters: * Form classes setup and usage - Form classes, validation, MapStruct mappers, and service layers * HTML Template Integration - Thymeleaf components, validation classes, and UI integration * [entity-checkers-setup-and-usage] - Entity Checkers for event-driven business rules
5. Data model setup
This section describes how to set up and manage the data model using SQL file autogeneration and Flyway migrations.
5.1. SQL File Autogeneration
The platform automatically generates SQL files based on your entity definitions. This process creates the necessary database schema files that can be used with Flyway for database migrations.
Automatic Generation Process
After the class is configured, run the application. The system will analyze changes in the data model of Java classes and generate an SQL script with the necessary changes V241012192920.sql in the project’s home directory (application.home=path), in the subdirectory hbm2ddl.
The generation process works as follows:
-
Entity Analysis: The system scans all JPA entity classes for changes
-
Schema Comparison: Compares current entity definitions with the existing database schema
-
SQL Generation: Creates appropriate DDL statements (CREATE, ALTER, DROP) for detected changes
-
File Creation: Generates timestamped migration files in the
hbm2ddldirectory -
Migration Integration: Files can be moved to Flyway migration directory for deployment
Entity Definition Example
Let’s look at the Participant entity as an example:
@Entity
@Table
@Audited
public class Participant extends AbstractAuditable implements NamedEntity, GithubDataSourceSubject, HasDocuments,
ProcessEntity, DocusignSigner, HasPendingDecisions, HasMetric {
public static final String DECISION_OWNER_TYPE = "PARTICIPANT";
@Column(nullable = false)
@Enumerated(value = EnumType.STRING)
private ParticipantStatus status = ParticipantStatus.NEW;
@ElementCollection(fetch = FetchType.EAGER)
@Enumerated(EnumType.STRING)
private Set<ParticipantRole> roles;
@ManyToOne(fetch = EAGER)
@JoinColumn(nullable = false, updatable = false)
private Application application;
@ManyToOne(fetch = EAGER)
@JoinColumn(nullable = false, updatable = false)
private Client client;
@Column(nullable = false)
@Enumerated(STRING)
private Employment employment;
@Column(nullable = false)
@Enumerated(STRING)
private Periodicity howOftenIncomeIsPaid;
@Embedded
private MonetaryAmount totalAnnualIncome;
@Embedded
private MonetaryAmount monthlyOutgoings;
@Column
private String githubUsername;
// getters and setters...
}
Enum Definitions
The entity uses several enums that define the possible values:
public enum ParticipantStatus implements InEnum<ParticipantStatus> {
NEW,
IN_PROCESS,
MANUAL_APPROVAL,
APPROVED,
DECLINED,
VOID;
public boolean isActive() {
return !this.in(DECLINED, VOID);
}
}
public enum Employment {
EMPLOYED,
HOMEMAKER,
UNEMPLOYED,
RETIRED,
SELF_EMPLOYED
}
public enum Periodicity {
MONTHLY,
FORTNIGHTLY,
WEEKLY,
UNDEFINED
}
5.2. Flyway Migration Integration
Migration File Structure
Flyway migration files are stored in the src/main/resources/db/migration/ directory and follow the naming convention:
V{version}__{description}.sql
For example:
V250530170222__init.sql V250609220043__participantStatus.sql
Generated SQL Example
Based on the Participant entity definition, the system generates the following SQL:
create table participant (
id uuid not null,
created_at timestamp(6) with time zone not null,
updated_at timestamp(6) with time zone not null,
employment varchar(255) not null,
how_often_income_is_paid varchar(255) not null,
monthly_outgoings_currency varchar(3),
monthly_outgoings_number numeric(19,2),
total_annual_income_currency varchar(3),
total_annual_income_number numeric(19,2),
created_by uuid,
updated_by uuid,
-- Foreign key to application table
application_id uuid not null,
-- Foreign key to client table
client_id uuid not null,
primary key (id)
);
-- Foreign key constraints for participant table
-- Links participant to their associated loan application
alter table if exists participant
add constraint FKa8akyngsbkcpy4ev19q53x56h
foreign key (application_id)
references application;
-- Links participant to their client profile containing personal information
alter table if exists participant
add constraint FKcmejtugfqk653qthh0jalsx54
foreign key (client_id)
references client;
Migration Workflow
-
Entity Definition: Define your entity classes with appropriate JPA annotations
-
Application Execution: Run the application to trigger the automatic analysis process
-
SQL Autogeneration: The platform analyzes entity changes and generates SQL scripts in the
hbm2ddlsubdirectory -
Migration File Preparation: Move generated SQL files from
hbm2ddlto the Flyway migration directory (src/main/resources/db/migration/) -
File Naming: Rename files to follow Flyway convention:
V{version}__{description}.sql -
Flyway Execution: During application startup, Flyway executes pending migrations in version order
-
Schema Versioning: Database schema version is tracked automatically in the
schema_versiontable
Best Practices
-
Incremental Changes: Create separate migration files for each schema change
-
Descriptive Names: Use clear, descriptive names for migration files
-
Testing: Test migrations on development environments before production
-
Rollback Strategy: Consider rollback scenarios when designing schema changes
Migration File Example
Here’s an actual migration file that adds participant status functionality:
-- Migration: Add participant status functionality
-- Add status column to audit table (for historical tracking)
alter table if exists aud_participant
add column status varchar(255);
-- Add status column to main participant table
alter table if exists participant
add column status varchar(255);
-- Set default status for all existing participants
update participant set status = 'NEW';
-- Make status column mandatory after setting default values
alter table if exists participant
alter column status set not null;
This approach ensures that your database schema evolves in a controlled, versioned manner while maintaining data integrity throughout the development lifecycle.
5.3. Troubleshooting Data Model Issues
SQL Generation Problems
No SQL Files Generated
Problem: Need to generate SQL migration files but no files appear in hbm2ddl directory
Solution:
-
Check application.home property:
# In application.properties application.home=/path/to/your/project # Or use relative path application.home=. -
Verify entities are detected:
// Ensure entities have proper annotations @Entity @Table(name = "your_table_name") public class YourEntity extends AbstractAuditable<UUID> { } -
Check entity scanning configuration:
// Ensure entities are in scanned packages @EntityScan(basePackages = {"com.yourpackage.entity"}) @SpringBootApplication public class Application { }
Flyway Migration Issues
Wrong Migration Order
Problem: Migrations execute in wrong order due to versioning issues
Solution:
-
Check version numbering:
# Correct format: VyyyyMMddHHmmss__description.sql V250610120000__add_participant_status.sql V250610130000__add_participant_employment.sql # Wrong (will execute in wrong order): V1__add_status.sql V10__add_employment.sql # This executes before V2! -
Use consistent timestamp format:
# Generate timestamp for new migration date +"%y%m%d%H%M%S" # Use this in filename: V250610143022__your_description.sql -
Fix ordering with new migration:
-- If wrong order was applied, create corrective migration -- V250610150000__correct_previous_changes.sql
Database Connection Issues
Connection Pool Exhaustion
Problem: HikariPool: Connection is not available errors
Solution:
-
Check connection pool settings:
# In application.properties spring.datasource.hikari.maximum-pool-size=20 spring.datasource.hikari.minimum-idle=5 spring.datasource.hikari.connection-timeout=20000 spring.datasource.hikari.idle-timeout=300000 -
Monitor connection usage:
# Enable connection pool metrics spring.datasource.hikari.register-mbeans=true management.endpoints.web.exposure.include=metrics -
Check for connection leaks:
// Ensure @Transactional is used properly @Service @Transactional // This ensures connections are properly closed public class YourService { }
Database Lock Issues
Problem: Migrations hang or fail with lock timeout errors
Solution:
-
Check for long-running transactions:
-- PostgreSQL: Find blocking queries SELECT pid, usename, application_name, state, query FROM pg_stat_activity WHERE state != 'idle' ORDER BY query_start; -
Kill blocking sessions (carefully):
-- Terminate specific session SELECT pg_terminate_backend(12345); -- Replace with actual PID
Performance Issues
Slow Entity Loading
Problem: Entity queries are slow or cause N+1 query problems
Solution:
-
Add database indexes using Hibernate annotations:
@Entity @Table(name = "participants", indexes = { @Index(name = "idx_participants_status", columnList = "status"), @Index(name = "idx_participants_employment", columnList = "employment"), @Index(name = "idx_participants_application_id", columnList = "application_id") }) public class Participant extends AbstractAuditable<UUID> { @Enumerated(EnumType.STRING) @Column(name = "status") private ParticipantStatus status; @Enumerated(EnumType.STRING) @Column(name = "employment") private Employment employment; @ManyToOne(fetch = FetchType.LAZY) @JoinColumn(name = "application_id") private Application application; }Alternative: Add indexes via migration file:
-- V250610120000__add_performance_indexes.sql CREATE INDEX idx_participants_status ON participants(status); CREATE INDEX idx_participants_employment ON participants(employment); CREATE INDEX idx_participants_application_id ON participants(application_id); -
Use proper fetch strategies:
@Entity public class Participant { @ManyToOne(fetch = FetchType.LAZY) // Don't use EAGER unless necessary private Application application; @OneToMany(fetch = FetchType.LAZY, mappedBy = "participant") private List<Document> documents; } -
Enable query logging to diagnose:
# In application.properties (development only) spring.jpa.show-sql=true spring.jpa.properties.hibernate.format_sql=true logging.level.org.hibernate.SQL=DEBUG
6. Form classes setup and usage
This section describes how to set up and manage form classes for data input validation and processing in the application.
6.1. Form Class Architecture
The platform uses form classes to handle user input validation, data binding, and form processing. Form classes serve as DTOs (Data Transfer Objects) that define the structure and validation rules for user interfaces.
Form Class Hierarchy
The application uses a hierarchical form structure:
-
Main Forms: Top-level forms like
ClientFormandApplicationForm -
Nested Forms: Component forms like
IndividualInfoFormandContactInfoForm -
Validation: Bean Validation (JSR-303) annotations for field validation
Form Class Examples
ClientForm Structure
The ClientForm class handles client registration and profile management:
@Valid
@NotNull
private IndividualInfoForm individualInfo;
@Valid
@NotNull
private ContactInfoForm contactInfo;
ApplicationForm Structure
The ApplicationForm class manages loan application data:
@Valid
private ParticipantForm borrowerParticipant;
Nested Form Components
Personal information component:
@NotBlank
private String nationalId;
@NotBlank
private String fullName;
@PastOrPresent
@DateTimeFormat(pattern = ValidationUtils.PATTERN_DATEPICKER_FORMAT)
private LocalDate dateOfBirth;
@NotNull
private Country residenceCountry;
Validation Annotations Used
The form classes use standard Bean Validation (JSR-303) annotations:
@NotNull // Field cannot be null
@NotBlank // String field cannot be null, empty, or whitespace only
@Email // Valid email format
@PastOrPresent // Date must be in the past or present
@Valid // Cascade validation to nested objects
@Phone // Custom phone validation (platform-specific)
6.2. Form Processing Architecture
Action Classes
The platform uses generic action classes to handle form operations:
@Controller
public class CreateClientAction extends EntityCreateController<UUID, Client, ClientForm> {
@Override
protected boolean isOwnPage() {
return false;
}
}
@Controller
public class EditClientAction extends EditEntityActionController<UUID, Client, ClientForm> {
}
These actions are parameterized with:
-
ID Type:
UUID- The entity identifier type -
Entity Type:
Client- The JPA entity class -
Form Type:
ClientForm- The form DTO class
Form Service Layer
Actions delegate form processing to specialized service classes:
@Service
public class ClientFormService extends EntityFormService<Client, ClientForm, UUID> {
The EntityFormService provides:
-
Entity to Form mapping: Converting entities to form objects for editing
-
Form to Entity mapping: Converting form submissions to entity objects
-
Validation integration: Coordinating with Bean Validation
-
Persistence operations: Saving and updating entities
MapStruct Mappers
Form-to-entity conversion is handled by MapStruct mappers:
@Mapper
public interface ClientFormMapper extends EntityToFormMapper<Client, ClientForm> {
@Mapper(uses = ParticipantFormMapper.class)
public interface ApplicationFormMapper extends EntityToFormMapper<Application, ApplicationForm> {
@Mapper
public interface ParticipantFormMapper extends EntityToFormMapper<Participant, ParticipantForm> {
MapStruct automatically generates implementation classes that provide:
-
Bidirectional mapping: Entity ↔ Form conversion
-
Nested object mapping: Automatic handling of complex object structures
-
Type conversion: Automatic conversion between compatible types
-
Null handling: Safe mapping of optional fields
For detailed information about MapStruct features and configuration, see the official MapStruct documentation.
Processing Flow
The complete form processing flow:
-
Action Invocation:
CreateClientActionorEditClientActionis called -
Service Delegation: Action delegates to
ClientFormService -
Mapper Usage: Service uses
ClientFormMapperfor conversions -
Entity Operations: Service performs database operations
-
Response Generation: Converted data is returned to the controller
EditClientAction<UUID, Client, ClientForm>
↓
ClientFormService.prepareEditModel(UUID id)
↓
ClientFormMapper.entityToForm(Client entity)
↓
ClientForm (ready for template rendering)
CreateClientAction<UUID, Client, ClientForm>
↓
ClientFormService.save(ClientForm form)
↓
ClientFormMapper.formToEntity(ClientForm form)
↓
Client entity (persisted to database)
6.3. Template Integration
Form classes integrate with HTML templates using Thymeleaf for rendering user interfaces. The templates use nested field access (dot notation) and reusable form components for consistent styling and validation.
For detailed information about HTML template integration, form components, and Thymeleaf usage, see HTML Template Integration.
6.4. Troubleshooting Form Issues
Form Validation Problems
Server-Side Validation Not Working
Problem: Invalid data reaches the service layer despite validation annotations
Solution:
-
Check @Valid annotations are present:
// On controller method parameters @PostMapping public String save(@Valid @ModelAttribute ClientForm form, BindingResult result) { if (result.hasErrors()) { return "client/edit"; } // ... } -
Verify nested object validation:
public class ClientForm { @Valid // This is crucial for nested validation @NotNull private IndividualInfoForm individualInfo; @Valid // Don't forget this @NotNull private ContactInfoForm contactInfo; } -
Check validation annotations are correct:
public class IndividualInfoForm { @NotBlank(message = "Full name is required") @Size(max = 255, message = "Full name cannot exceed 255 characters") private String fullName; @Email(message = "Please provide a valid email address") private String email; }
MapStruct Mapping Issues
Compilation Errors
Problem: No property named "X" exists in source parameter or similar MapStruct errors
Solution:
-
Verify property names match exactly:
// Entity public class Client { private IndividualInfo individualInfo; // Property name public IndividualInfo getIndividualInfo() { return individualInfo; } } // Form - property name must match public class ClientForm { private IndividualInfoForm individualInfo; // Same name public IndividualInfoForm getIndividualInfo() { return individualInfo; } } -
Check nested object mapping:
@Mapper public interface ClientFormMapper extends EntityToFormMapper<Client, ClientForm> { // Explicit mapping may be needed for complex cases @Mapping(source = "individualInfo.fullName", target = "individualInfo.fullName") @Mapping(source = "contactInfo.email", target = "contactInfo.email") ClientForm entityToForm(Client entity); } -
Rebuild after changes:
# MapStruct generates code at compile time mvn clean compile # Check generated classes in target/generated-sources/annotations/
Form Templates
Form Template Structure
The form template begins with a <form> tag. The following required blocks are expected within the form:
The form__actions block contains form control elements:
Form title (<h4>) - displayed only when the form is opened as a separate page. When the form is displayed in a modal window, the title is automatically hidden Button block - contains form action buttons (save, cancel, etc.) with ml-auto class for alignment
| The location of the form__actions block is not important - it can be placed either at the beginning or at the end of the <form> element. |
<div class="form__actions">
<h4>Form Title</h4>
<div class="ml-auto">
<!-- Form buttons -->
</div>
</div>
The form-body block is the main container for form fields and is responsible for the grid layout. The following elements are placed inside this block:
form-group - form field groups
app_toggled - togglable blocks
<h4> - section subheadings
<div class="form-body">
<h4>Form Section</h4>
<div class="form-group">
<!-- Form fields -->
</div>
<div class="app_toggled">
<!-- Togglable content -->
</div>
</div>
Additional Elements
All additional blocks (tables, information blocks) are inserted directly into the <form> element, outside the form-body block. IMPORTANT: Tables must be wrapped in a <div>:
<form>
<div class="form__actions">
<!-- ... -->
</div>
<div class="form-body">
<!-- ... -->
</div>
<div>
<table>
<!-- Table content -->
</table>
</div>
</form>
Two-Column Layout for Modal Forms
The form-2-column class is used to add a two-column layout to forms in modal windows. This class is added to the form-body block.
The two-column layout uses CSS Grid with the following characteristics:
Grid columns: Two equal columns (1fr 1fr) Column gap: 60px spacing between columns Width: Full width of the container (100%)
form-group elements: Automatically placed in columns (alternating left/right) app_toggled elements: Also placed in columns like form-group elements All other elements: Span both columns (full width) using grid-column: 1 / -1
The form__container class can be used within form-2-column to create a nested two-column grid:
Spans both columns of the parent grid Creates its own two-column grid with the same 60px gap Useful for grouping related fields that need their own column layout
<div class="form-body form-2-column">
<h4>Section Title</h4> <!-- Spans both columns -->
<div class="form-group">
<!-- Placed in left column -->
</div>
<div class="form-group">
<!-- Placed in right column -->
</div>
<div class="app_toggled">
<!-- Placed in next available column (not spanning full width) -->
</div>
<div class="form__container">
<!-- Nested two-column grid -->
<div class="form-group"><!-- Left column --></div>
<div class="form-group"><!-- Right column --></div>
</div>
</div>
.form-2-column {
display: grid;
grid-template-columns: 1fr 1fr;
grid-column-gap: 60px;
width: 100%;
*:not(.form-group):not(.app-toggled) {
grid-column: 1 / -1;
}
.form__container {
grid-column: 1 / -1;
display: grid;
grid-template-columns: 1fr 1fr;
grid-column-gap: 60px;
}
}
| Both form-group and app_toggled elements are automatically distributed across columns. All other elements (headings, tables, containers) will span the full width of both columns. |
Complete Structure Example
<form>
<!-- Action block (can be at the beginning or end) -->
<div class="form__actions">
<h4>Form Name</h4>
<div class="ml-auto">
<button type="submit">Save</button>
<button type="button">Cancel</button>
</div>
</div>
<!-- Main form body (standard layout) -->
<div class="form-body">
<h4>Basic Information</h4>
<div class="form-group">
<label>Field 1</label>
<input type="text" />
</div>
<div class="app_toggled">
<div class="form-group">
<label>Field 2</label>
<input type="text" />
</div>
</div>
</div>
<!-- Main form body (two-column layout for modals) -->
<div class="form-body form-2-column">
<h4>Modal Form Section</h4> <!-- Spans both columns -->
<div class="form-group">
<label>Field 1</label>
<input type="text" />
<!-- Left column -->
</div>
<div class="form-group">
<label>Field 2</label>
<input type="text" />
<!-- Right column -->
</div>
<div class="app_toggled">
<div class="form-group">
<label>Field 3</label>
<input type="text" />
</div>
</div>
<div class="form__container">
<!-- Nested two-column grid spanning full width -->
<div class="form-group">
<label>Nested Field 1</label>
<input type="text" />
</div>
<div class="form-group">
<label>Nested Field 2</label>
<input type="text" />
</div>
</div>
</div>
<!-- Additional elements -->
<div>
<table>
<thead>
<tr>
<th>Column 1</th>
<th>Column 2</th>
</tr>
</thead>
<tbody>
<!-- Table data -->
</tbody>
</table>
</div>
</form>
7. HTML Template Integration
This section describes how form classes integrate with HTML templates using Thymeleaf for rendering user interfaces.
7.1. Template Structure
The application uses Thymeleaf templates to render forms with automatic data binding and validation integration.
Client Form Template
The client edit form demonstrates nested form structure:
<h2 class="form-group__title" th:text="#{client.clientInfo}">Personal
Information</h2>
<th:block
th:insert="~{/form/components :: text(#{client.individualInfo.fullName},
'individualInfo.fullName', 'v-required v-name')}"
th:with="maxlength = 120" />
<th:block
th:insert="~{/form/components :: text(#{client.individualInfo.nationalId},
'individualInfo.nationalId', 'v-required')}" />
<th:block
th:insert="~{/form/components :: date (#{client.individualInfo.birthDate},
'individualInfo.dateOfBirth', '')}"
th:with="maxDate = ${#dates.createNow()}" />
<th:block
th:insert="~{/form/components :: select(#{client.address.stateOfResidence},
'individualInfo.residenceCountry', ${countries})}" />
<h2 class="form-group__title" th:text="#{client.contactInfo}">Contact
Information</h2>
<th:block
th:insert="~{/form/components :: text(#{client.contactInfo.email},
'contactInfo.email', 'v-required v-email')}" />
<th:block
th:insert="~{/form/components :: text(#{client.contactInfo.phone},
'contactInfo.phone', 'v-required v-phone')}" />
Key features:
-
Nested field access: Uses dot notation like
individualInfo.fullName -
Validation classes: CSS classes for client-side validation (
v-required,v-email) -
Component reuse: Uses Thymeleaf fragments for consistent field rendering
Application Form Template
The application edit form shows financial data handling:
<h2 th:text="#{application.borrowerInfo}">Borrower
Information</h2>
<th:block
th:insert="~{/form/components :: select(#{participant.employment},
'borrowerParticipant.employment', ${employmentTypes})}" />
<th:block
th:insert="~{/form/components :: select(#{participant.howOftenIncomeIsPaid},
'borrowerParticipant.howOftenIncomeIsPaid', ${periodicities})}" />
<h2 class="mt-10" th:text="#{participant.financialInfo}">Financial
Information</h2>
<th:block
th:insert="~{/form/components :: amount(#{participant.totalAnnualIncome},
'borrowerParticipant.totalAnnualIncome', 'v-required')}" />
<th:block
th:insert="~{/form/components :: amount(#{participant.monthlyOutgoings},
'borrowerParticipant.monthlyOutgoings', '')}" />
Features:
-
Enum handling: Select dropdowns for
employmentandperiodicities -
Monetary amounts: Special
amountcomponent for financial fields -
Nested participant: Access to
borrowerParticipantfields
7.2. Form Component System
The platform uses Thymeleaf fragments for consistent form rendering across all forms. These components are defined in /form/components.html and provide standardized UI elements with built-in validation support.
Available Form Components
Text Input Component
~{/form/components :: text(name, fieldname, inputclass)}
| Parameter | Type | Description |
|---|---|---|
|
String (i18n key) |
Label text for the input field (e.g., |
|
String |
Field path for data binding (e.g., |
|
String |
CSS validation classes (e.g., |
-
maxlength- Maximum character limit (default: 256) -
minlength- Minimum character limit (default: 0) -
placeholder- Placeholder text for the input field
<th:block th:insert="~{/form/components :: text(
#{client.individualInfo.fullName},
'individualInfo.fullName',
'v-required v-armenian-name')}"
th:with="maxlength = 120, placeholder = #{placeholder.fullName}" />
Select Dropdown Component
~{/form/components :: select(name, fieldname, values)}
| Parameter | Type | Description |
|---|---|---|
|
String (i18n key) |
Label text for the select field |
|
String |
Field path for data binding |
|
Collection/Map |
Options data (Map for key-value pairs, Collection for simple lists) |
<th:block th:insert="~{/form/components :: select(
#{client.address.stateOfResidence},
'individualInfo.residenceCountry',
${countries})}" />
Date Picker Component
~{/form/components :: date(name, fieldname, inputclass)}
| Parameter | Type | Description |
|---|---|---|
|
String (i18n key) |
Label text for the date field |
|
String |
Field path for data binding |
|
String |
CSS validation classes (optional) |
-
maxDate- Maximum selectable date -
minDate- Minimum selectable date -
startDate- Default selected date
<th:block th:insert="~{/form/components :: date(
#{client.individualInfo.birthDate},
'individualInfo.dateOfBirth',
'v-required')}"
th:with="maxDate = ${#dates.format(#dates.createNow())}" />
Amount/Currency Component
~{/form/components :: amount(name, fieldname, inputclass)}
| Parameter | Type | Description |
|---|---|---|
|
String (i18n key) |
Label text for the amount field |
|
String |
Field path for data binding |
|
String |
CSS validation classes (e.g., |
-
inputAmountPrefix- Prefix for field IDs (optional) -
currencies- Available currency options
<th:block th:insert="~{/form/components :: amount(
#{participant.totalAnnualIncome},
'borrowerParticipant.totalAnnualIncome',
'v-required v-armenian-tax-id')}" />
Checkbox Component
~{/form/components :: checkbox(name, fieldname, inputclass)}
| Parameter | Type | Description |
|---|---|---|
|
String (i18n key) |
Label text for the checkbox |
|
String |
Field path for data binding |
|
String |
CSS classes for styling/validation |
<th:block th:insert="~{/form/components :: checkbox(
#{client.agreeToTerms},
'agreeToTerms',
'v-required')}" />
Textarea Component
~{/form/components :: textarea(name, fieldname, inputclass)}
| Parameter | Type | Description |
|---|---|---|
|
String (i18n key) |
Label text for the textarea |
|
String |
Field path for data binding |
|
String |
CSS validation classes |
-
rows- Number of textarea rows (default: 5) -
maxlength- Maximum character limit
<th:block th:insert="~{/form/components :: textarea(
#{application.comments},
'comments',
'v-required')}"
th:with="rows = 3, maxlength = 500" />
Radio Button Component
~{/form/components :: radio(name, fieldname, params)}
| Parameter | Type | Description |
|---|---|---|
|
String (i18n key) |
Label text for the radio group |
|
String |
Field path for data binding |
|
Map |
Key-value pairs for radio options |
<th:block th:insert="~{/form/components :: radio(
#{client.gender},
'gender',
${genderOptions})}" />
File Upload Component
~{/form/components :: fileInput(name, filelabel, fieldname, inputclass)}
| Parameter | Type | Description |
|---|---|---|
|
String (i18n key) |
Label text for the file input |
|
String |
Button text for file selection |
|
String |
Field path for data binding |
|
String |
CSS classes for styling/validation |
<th:block th:insert="~{/form/components :: fileInput(
#{document.upload},
#{button.chooseFile},
'documentFile',
'v-required')}" />
Read-only Component
~{/form/components :: readonly(name, fieldname, inputclass)}
| Parameter | Type | Description |
|---|---|---|
|
String (i18n key) |
Label text for the read-only field |
|
String |
Field path for data binding |
|
String |
CSS classes for styling |
<th:block th:insert="~{/form/components :: readonly(
#{client.id},
'id',
'')}" />
Component Benefits
This component system ensures:
-
Consistency: All forms use the same styling and behavior
-
Maintainability: Changes to form components affect all forms
-
Validation Integration: Client-side validation works seamlessly
-
Accessibility: Standard form components ensure accessibility compliance
-
Internationalization: Built-in support for i18n message keys
-
Reusability: Components can be used across different forms and contexts
7.3. Client-Side Validation Classes
The platform provides CSS-based validation classes that integrate with jQuery Validation for client-side form validation:
Standard Validation Classes
| CSS Class | Description | Usage Example |
|---|---|---|
|
Field is mandatory and cannot be empty |
|
|
Field must contain a valid number |
|
|
Field must contain only digits (0-9) |
|
|
Field must contain a valid email address |
|
|
Field must contain a valid URL |
|
|
Field must contain a valid phone number |
|
|
Field must contain a positive number (> 0) |
|
|
Field must contain valid name characters (letters, spaces, hyphens, apostrophes), max 256 characters |
|
Custom Validation Methods
The platform extends jQuery Validation with custom validation methods:
// Armenian name validation (Armenian letters, spaces, hyphens only)
$.validator.addMethod('armenianName', function(value, element) {
const ARMENIAN_NAME_REGEX = /^[\u0531-\u0556\u0561-\u0587\s\-']+$/;
return this.optional(element) || ARMENIAN_NAME_REGEX.test(value);
});
// Tax identification number validation (Armenian format)
$.validator.addMethod('armenianTaxId', function(value, element) {
const TAX_ID_REGEX = /^\d{8}$/;
return this.optional(element) || TAX_ID_REGEX.test(value);
});
// Armenian postal code validation
$.validator.addMethod('armenianPostal', function(value, element) {
const POSTAL_REGEX = /^\d{4}$/;
return this.optional(element) || POSTAL_REGEX.test(value);
});
Validation Class Rules Mapping
The CSS classes are mapped to validation rules using jQuery Validation:
$.validator.addClassRules({
'v-armenian-name': {armenianName: true, maxlength: 256},
'v-armenian-tax-id': {armenianTaxId: true},
'v-armenian-postal': {armenianPostal: true},
});
Usage in Templates
Validation classes are applied as the third parameter in form component calls:
<!-- Required text field with name validation -->
<th:block th:insert="~{/form/components :: text(
#{client.individualInfo.fullName},
'individualInfo.fullName',
'v-required v-name')}" />
<!-- Required email field -->
<th:block th:insert="~{/form/components :: text(
#{client.contactInfo.email},
'contactInfo.email',
'v-required v-email')}" />
<!-- Required positive amount field -->
<th:block th:insert="~{/form/components :: amount(
#{participant.totalAnnualIncome},
'borrowerParticipant.totalAnnualIncome',
'v-required v-positive')}" />
Combining Validation Classes
Multiple validation classes can be combined using space separation:
-
'v-required v-email'- Required email field -
'v-required v-name'- Required name field with character validation -
'v-required v-positive'- Required positive number field -
'v-number v-positive'- Optional positive number field
8. Fast Search System
Add search functionality to your entities with minimal configuration.
8.1. What You Need to Know
Fast Search lets users search across all your application entities. It’s already enabled and integrated into the UI - you just need to configure which entities are searchable.
8.2. Configuration
8.3. Quick Setup
1. Make Entity Searchable
Add @Indexed to your entity and mark searchable fields:
@Entity
@Indexed
public class Client extends AbstractAuditable<UUID> {
@FullTextField // Natural language search
private String fullName;
@KeywordField // Exact matches only
private String nationalId;
@IndexedEmbedded // Include nested fields
private ContactInfo contactInfo;
}
2. Configure Search Entity
Create a configuration class:
@Configuration
@ConditionalOnProperty("hibernate.search.enabled")
public class ClientSearchConfig {
@Bean
FastSearchEntity<Client> clientSearchEntity() {
return new FastSearchEntity<>(Client.class, Client::getDisplayedName);
}
}
That’s it! Search is now available in the application UI at /search.
8.4. Field Types
@FullTextField - For names, descriptions, natural language text
@KeywordField - For IDs, codes, statuses, exact matches
@IndexedEmbedded - Include fields from nested objects
8.5. FastSearchEntity Configuration
The FastSearchEntity bean controls how your entity appears in search results.
Basic Configuration
@Bean
FastSearchEntity<Client> clientSearchEntity() {
return new FastSearchEntity<>(Client.class, Client::getDisplayedName);
}
Parameters:
-
Client.class- The entity class to make searchable -
Client::getDisplayedName- Function that returns the display name for search results
Custom Display Names
Control how search results are displayed:
@Bean
FastSearchEntity<Client> clientSearchEntity() {
return new FastSearchEntity<>(Client.class, client -> {
String name = client.getIndividualInfo().getFullName();
String id = client.getIndividualInfo().getNationalId();
return name + " (" + id + ")";
});
}
Layout Templates
Specify custom result templates:
@Bean
FastSearchEntity<Client> clientSearchEntity() {
FastSearchEntity<Client> entity = new FastSearchEntity<>(Client.class, Client::getDisplayedName);
entity.setLayoutTemplateName("/client/search :: client-result");
return entity;
}
This uses /client/search :: client-result template for displaying this entity in search results. Default template is /form/search :: default.
Multiple Entities
Configure multiple entities in one config class:
@Configuration
@ConditionalOnProperty("hibernate.search.enabled")
public class SearchConfig {
@Bean
FastSearchEntity<Client> clientSearchEntity() {
return new FastSearchEntity<>(Client.class, Client::getDisplayedName);
}
@Bean
FastSearchEntity<Application> applicationSearchEntity() {
return new FastSearchEntity<>(Application.class, app ->
"Application #" + app.getId());
}
}
8.6. Entity Labels
Add entity type labels to messages.properties for proper display:
entity.client=Client
entity.application=Application
entity.participant=Participant
8.7. Index Management
Search indexes are created automatically when the application starts.
8.8. Search Examples
Users can search with:
john smith # Find both terms
"john smith" # Exact phrase
fullName:john # Search specific field
john* # Wildcard search
john AND smith # Boolean operators
9. Entity Counters and Launchpad Integration
Create dashboard counters that display pending decisions and workflow statistics, providing seamless integration with the Launchpad decision management system.
9.1. What You’ll Learn
After reading this chapter, you’ll know how to:
-
Create entity counters that integrate with Launchpad pending decisions
-
Track workflow states and decision progress through counters
-
Implement counters for entities with pending decision workflows
-
Navigate from dashboard counters directly to Launchpad decision interfaces
-
Configure counter-to-Launchpad routing for efficient workflow management
-
Test and troubleshoot counter-Launchpad integrations
9.2. The Big Picture
Entity counters serve as workflow command centers that bridge dashboard visibility with Launchpad decision-making:
-
Decision Visibility: Show counts of entities with pending decisions
-
Workflow Progress: Track entities at different workflow stages
-
Quick Access: Click counters to navigate directly to Launchpad decision interfaces
-
Priority Management: Surface entities requiring immediate attention
-
Business Intelligence: Monitor decision throughput and bottlenecks
Think of counters as workflow entry points - they identify entities needing decisions and provide direct access to the Launchpad where those decisions are made.
Counters + Launchpad Workflow
The integration works as follows:
-
Counter Detection: Counter identifies entities with pending decisions or specific workflow states
-
Dashboard Display: Shows count and entity summary information
-
Launchpad Navigation: Clicking entities opens them in Launchpad for decision-making
-
Decision Processing: Users make decisions through Launchpad interface
-
Counter Updates: Counters automatically reflect workflow progress
9.3. Core Concepts
EntityCounter Interface
The EntityCounter interface defines how to identify and count entities:
public interface EntityCounter<T> {
boolean isEntityMarked(T entity);
String getName();
}
Key Methods:
-
isEntityMarked(T entity): Returnstrueif entity should be counted -
getName(): Returns counter identifier used for templates and internationalization
Counter Architecture
Counters work by:
1. Entity Evaluation: Check each entity against counter criteria
2. Count Aggregation: Platform counts entities where isEntityMarked() returns true
3. Dashboard Display: Show count and provide link to filtered entity list
4. Navigation: Users click to see entities that match the counter
9.4. Implementation Examples
Simple Status-Based Counter
The ApplicationManualReviewCounter counts applications requiring manual review:
package com.timvero.example.admin.application.counter;
import com.timvero.example.admin.application.entity.Application;
import com.timvero.example.admin.application.entity.ApplicationStatus;
import com.timvero.ground.entity_marker.counter.EntityCounter;
import org.springframework.core.annotation.Order;
import org.springframework.stereotype.Component;
@Component
@Order(2000)
public class ApplicationManualReviewCounter implements EntityCounter<Application> {
@Override
public boolean isEntityMarked(Application entity) {
return entity.getStatus() == ApplicationStatus.MANUAL_REVIEW;
}
@Override
public String getName() {
return "applicationManualReview";
}
}
Key Implementation Details:
-
@Component: Registers counter with Spring container
-
@Order(2000): Controls display order on dashboard (lower numbers appear first)
-
Status Check: Simple enum comparison for entity filtering
-
Counter Name:
"applicationManualReview"used for templates and i18n
Label-Based Counter
The CreditActiveAndPaidCounter leverages entity labels for complex business logic:
package com.timvero.example.admin.credit.counter;
import com.timvero.example.admin.credit.entity.ExampleCredit;
import com.timvero.example.admin.credit.label.CreditPaidLabel;
import com.timvero.ground.entity_marker.counter.EntityCounter;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.core.annotation.Order;
import org.springframework.stereotype.Component;
@Component
@Order(3000)
public class CreditActiveAndPaidCounter implements EntityCounter<ExampleCredit> {
@Autowired
private CreditPaidLabel creditPaidLabel;
@Override
public boolean isEntityMarked(ExampleCredit entity) {
return creditPaidLabel.isEntityMarked(entity);
}
@Override
public String getName() {
return "creditActiveAndPaid";
}
}
This pattern demonstrates:
-
Label Reuse: Leverages existing
CreditPaidLabellogic -
Dependency Injection: Uses
@Autowiredto access label logic -
Business Logic Delegation: Keeps counter simple by delegating to specialized label
The Supporting Label
The CreditPaidLabel contains the actual business logic:
package com.timvero.example.admin.credit.label;
import static com.timvero.example.admin.credit.CreditCalculationConfiguration.ACTIVE;
import com.timvero.example.admin.credit.entity.ExampleCredit;
import com.timvero.ground.entity_marker.label.EntityLabel;
import java.util.Optional;
import javax.money.MonetaryAmount;
import org.springframework.core.annotation.Order;
import org.springframework.stereotype.Component;
@Component
@Order(2000)
public class CreditPaidLabel implements EntityLabel<ExampleCredit> {
@Override
public boolean isEntityMarked(ExampleCredit entity) {
if (entity.getActualSnapshot() == null) {
return false;
}
Optional<MonetaryAmount> totalDebt = entity.getActualSnapshot().getDebt().getTotal();
return entity.getActualSnapshot().getStatus().equals(ACTIVE) && totalDebt.isEmpty();
}
@Override
public String getName() {
return "paid";
}
}
Complex Business Logic:
-
Null Safety: Checks for
actualSnapshotexistence -
Status Validation: Verifies credit is in
ACTIVEstatus -
Debt Calculation: Confirms total debt is empty (paid off)
-
Monetary Handling: Works with
MonetaryAmountfor financial calculations
9.5. Configuration and Setup
Counter Registration
Counters are automatically discovered by Spring component scanning:
@Component
@Order(1000) // Lower numbers = higher priority (appear first)
public class MyEntityCounter implements EntityCounter<MyEntity> {
@Override
public boolean isEntityMarked(MyEntity entity) {
// Your counting logic here
return entity.getStatus() == MyEntityStatus.NEEDS_ATTENTION;
}
@Override
public String getName() {
return "myEntityCounter"; // Used for templates and i18n
}
}
Internationalization Setup
Add counter labels to your messages file:
File: src/main/resources/messages_en.properties
counter.applicationManualReview=Applications for Manual Review
counter.creditActiveAndPaid=Active and Paid Loans
counter.myEntityCounter=My Custom Counter
Pattern: counter.{counterName}={Display Label}
Display Order Configuration
Use @Order annotation to control counter sequence on dashboard:
@Order(1000) // Appears first
public class HighPriorityCounter implements EntityCounter<Entity> { ... }
@Order(2000) // Appears second
public class MediumPriorityCounter implements EntityCounter<Entity> { ... }
@Order(3000) // Appears third
public class LowPriorityCounter implements EntityCounter<Entity> { ... }
Best Practice: Use increments of 1000 to allow for future insertions.
9.6. Dashboard Integration
Counter Templates
Entity counters use shared templates for consistent display:
Dashboard Counter Template (templates/dashboard/counter/application.html):
<!DOCTYPE html>
<html xmlns:th="http://www.thymeleaf.org">
<th:block th:fragment="entity(entities)">
<ul class="counter-list">
<li
class="counter-list__item"
th:each="entity,iterStat : ${entities}"
>
<button class="clickable counter-list__btn" th:href="@{${url} + '/' + ${entity.id}}" th:object="${entity}">
<p class="d-flex justify-content-between w-100">
<span class="text-left text-prompt w-75" th:text="*{displayedName}"></span>
<span th:text="*{#ldates.format(createdAt, 'MM/dd/yyyy')}"></span>
</p>
</button>
</li>
</ul>
</th:block>
</html>
Template Features:
-
Entity List: Displays entities matching the counter criteria
-
Clickable Items: Each entity links to detailed view
-
Date Display: Shows creation date for context
-
Consistent Styling: Uses platform CSS classes for unified appearance
Dashboard Layout Integration
The main dashboard template integrates all counters:
Dashboard View (templates/dashboard/view.html):
<div class="counters-wrapper">
<div class="counter__item" th:each="entity,iterStat : ${counterList}" th:object="${entity}">
<th:block th:insert="~{:: counterCard(${entity})}"/>
<section
th:data-for-counter="${entity}"
th:data-counter-url="'/counter/' + ${entity} + '?count=10'"
class="mt-5">
</section>
<button class="counter__item-btn js-show-full-counter-list" th:text="#{common.viewAll}"></button>
</div>
</div>
<th:block th:fragment="counterCard(counterName)">
<div class="counterCard">
<div class="counterCard-title">
<h4 th:text="#{'counter.' + ${counterName}}"></h4>
<div class="d-flex flex-column align-items-end">
<span class="d-flex counter">
<span
th:data-name="${counterName}"
th:data-title="#{'counter.' + ${counterName}}">
</span>
</span>
</div>
</div>
</div>
</th:block>
Dashboard Integration Points:
-
Counter Loop: Iterates through all registered counters
-
Dynamic URLs: Builds counter-specific endpoints
-
Internationalization: Uses counter names for i18n lookup
-
JavaScript Integration: Provides hooks for dynamic counter updates
9.7. Performance Considerations
Efficient Counting Logic
Avoid Heavy Operations: Keep isEntityMarked() logic lightweight
// Good: Simple property checks
public boolean isEntityMarked(Application app) {
return app.getStatus() == ApplicationStatus.MANUAL_REVIEW;
}
// Avoid: Complex database queries in counter logic
public boolean isEntityMarked(Application app) {
// Don't do this - too expensive for counting
List<Document> docs = documentRepository.findByApplicationId(app.getId());
return docs.stream().anyMatch(doc -> doc.getStatus() == DocumentStatus.MISSING);
}
Use Entity Labels: Delegate complex logic to specialized label classes
@Component
public class MyCounter implements EntityCounter<MyEntity> {
@Autowired
private MyComplexLabel complexLabel; // Reuse existing logic
@Override
public boolean isEntityMarked(MyEntity entity) {
return complexLabel.isEntityMarked(entity); // Delegate
}
}
Database Query Optimization
Entity Fetching: Ensure counter queries are optimized
// If your counter needs related entities, optimize the query
@Query("SELECT e FROM MyEntity e " +
"JOIN FETCH e.relatedEntity " +
"WHERE e.status = :status")
List<MyEntity> findEntitiesForCounter(@Param("status") EntityStatus status);
Pagination: Platform handles pagination automatically for counter views
9.8. Best Practices
Counter Design Guidelines
Single Responsibility: Each counter should track one specific business condition
// Good: Focused on one criteria
public class ApplicationManualReviewCounter implements EntityCounter<Application> {
public boolean isEntityMarked(Application app) {
return app.getStatus() == ApplicationStatus.MANUAL_REVIEW;
}
}
// Avoid: Multiple unrelated criteria
public class MixedCriteriaCounter implements EntityCounter<Application> {
public boolean isEntityMarked(Application app) {
return app.getStatus() == ApplicationStatus.MANUAL_REVIEW ||
app.getAmount().isGreaterThan(THRESHOLD) ||
app.getCreatedAt().isBefore(CUTOFF_DATE); // Too many different concerns
}
}
Meaningful Names: Use descriptive counter names for templates and i18n
// Good: Clear business meaning
public String getName() {
return "applicationManualReview"; // Clear what this counts
}
// Avoid: Generic or unclear names
public String getName() {
return "counter1"; // What does this count?
}
Consistent Ordering: Group related counters with logical order values
// Applications
@Order(1000) public class ApplicationSubmittedCounter { ... }
@Order(1100) public class ApplicationManualReviewCounter { ... }
@Order(1200) public class ApplicationApprovedCounter { ... }
// Credits
@Order(2000) public class CreditActiveCounter { ... }
@Order(2100) public class CreditOverdueCounter { ... }
@Order(2200) public class CreditPaidCounter { ... }
Internationalization Best Practices
Descriptive Labels: Make counter purposes clear to users
# Good: Clear business context
counter.applicationManualReview=Applications Requiring Review
counter.overduePayments=Overdue Payments
counter.incompleteDocumentation=Missing Documentation
# Avoid: Technical or vague labels
counter.status1=Status 1 Items
counter.items=Items
counter.counter=Counter
9.9. Troubleshooting
Common Issues
Counter Not Appearing on Dashboard: Counter component not registered
Solution: Ensure counter class is annotated with @Component and is in a
package scanned by Spring component scanning.
Counter Shows Zero Count: Logic never returns true
Solution: Debug the isEntityMarked() method with test data. Check that your
criteria match actual entity states in the database.
Counter Display Label Missing: No internationalization entry
Solution: Add counter.{counterName}=Display Label to messages_en.properties
file. Ensure the counter name matches exactly.
Counter Order Wrong: @Order values not set correctly
Solution: Use @Order annotation with appropriate numeric values. Lower
numbers appear first on the dashboard.
Performance Issues: Counter logic too expensive
Solution: Profile the isEntityMarked() method. Move complex logic to
entity labels or optimize database queries used by the counter.
Counter Updates Delayed: Changes don’t reflect immediately
Solution: Ensure workflow completion events properly trigger entity
updates. Consider adding @Transactional annotations with proper
propagation to avoid race conditions.
Entity counters provide powerful workflow command centers that seamlessly integrate dashboard visibility with Launchpad decision-making. They enable efficient workflow management by surfacing entities requiring attention and providing direct access to decision processing interfaces.
10. Launchpad Integration
Integrate your entities with the Launchpad decision-making system to provide users with unified workflow management and pending decision handling.
10.1. What You’ll Learn
After reading this chapter, you’ll know how to:
-
Integrate entities with the Launchpad system for decision management
-
Create custom decision entity extractors for your business entities
-
Configure Launchpad templates for entity-specific workflow display
-
Handle pending decision workflows through Launchpad interface
-
Provide contextual information for decision-making processes
10.2. The Big Picture
The Launchpad system provides a centralized interface where users can:
-
View and manage pending decisions across different entity types
-
Access entity-specific context and details for informed decision-making
-
Execute workflow actions directly from a unified dashboard
-
Track decision progress and status for multiple business entities
Think of Launchpad as a decision workbench - it aggregates all pending decisions from your application entities and presents them in a consistent, actionable interface.
10.3. Core Components
DecisionEntityExtractor
The DecisionEntityExtractor interface connects your entities to the Launchpad system. It provides the bridge between pending decisions and entity context.
Interface Definition
public interface DecisionEntityExtractor {
String getDecisionOwnerType();
LaunchpadEntityDto getByHolderId(Long holderId);
}
Key Methods:
-
getDecisionOwnerType(): Returns the entity type identifier for routing decisions -
getByHolderId(Long holderId): Retrieves entity context by pending decision holder ID
LaunchpadEntityDto
The LaunchpadEntityDto carries all necessary information for displaying entity context in Launchpad:
public class LaunchpadEntityDto {
private Long targetEntityId; // Primary entity ID
private Object parentEntity; // Parent/container entity
private String parentDisplayName; // Parent entity display name
private Object targetEntity; // Target entity with pending decisions
private String targetDisplayName; // Target entity display name
private String roleDescription; // Context description (roles, status, etc.)
private String templatePath; // Path to Launchpad template fragment
}
10.4. Implementation Example: Participant Launchpad Service
The ParticipantLaunchpadService demonstrates how to integrate participant entities with Launchpad:
Service Implementation
@Component
public class ParticipantLaunchpadService implements DecisionEntityExtractor {
@Autowired
private ParticipantRepository repository;
@Override
public String getDecisionOwnerType() {
return Participant.DECISION_OWNER_TYPE;
}
@Override
public LaunchpadEntityDto getByHolderId(Long holderId) {
Participant participant = repository.findByPendingDecisionHolderId(holderId)
.orElseThrow(() -> new RuntimeException("No target entity with holder id " + holderId));
Application application = participant.getApplication();
String roles = participant.getRoles().stream()
.map(role -> EnumUtils.getLocalizedValue(role, LocaleContextHolder.getLocale()))
.collect(Collectors.joining(", "));
return new LaunchpadEntityDto(participant.getId(), application, application.getDisplayedName(),
participant, participant.getDisplayedName(), roles,
"/participant/launchpad-body");
}
}
Key Implementation Details:
-
Decision Owner Type: Returns
Participant.DECISION_OWNER_TYPEto identify this extractor -
Entity Retrieval: Finds participant by pending decision holder ID
-
Context Building: Constructs display information including roles and parent application
-
Template Path: Specifies the template fragment path for rendering participant context
Entity Requirements
For entities to work with Launchpad, they must implement HasPendingDecisions:
@Entity
public class Participant extends AbstractAuditable<UUID> implements HasPendingDecisions {
public static final String DECISION_OWNER_TYPE = "PARTICIPANT";
@OneToOne(cascade = CascadeType.ALL)
@JoinColumn(nullable = false, updatable = false)
private PendingDecisionHolder pendingDecisionHolder =
new PendingDecisionHolder(DECISION_OWNER_TYPE);
@Override
public PendingDecisionHolder getPendingDecisionHolder() {
return pendingDecisionHolder;
}
}
10.5. Template Integration
Launchpad Template Fragment
Create a template fragment for your entity’s Launchpad display:
File: src/main/resources/templates/participant/launchpad-body.html
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml"
xmlns:th="http://www.thymeleaf.org">
<th:block th:fragment="details" th:object="${entityDto.targetEntity}">
<th:block th:insert="~{/launchpad/fragment/pending-decisions :: body(${entityDto})}"/>
<div class="mt-20 border-top">
<h4 class="mb-5 mt-5" th:text="#{participant.tab.details}"></h4>
<article
class="article pl-10 pr-10"
th:data-url="@{${@pathRegistry.getPath(entityDto.targetEntity) + '/tab/details'}}"
></article>
</div>
<div class="mt-20 border-top">
<h4 class="mb-5 mt-5" th:text="#{participant.tab.profile}"></h4>
<article
class="article"
th:data-url="@{${@pathRegistry.getPath(entityDto.targetEntity) + '/tab/profile'}}"
></article>
</div>
<div class="mt-20 border-top">
<h4 class="mb-5 mt-5" th:text="#{participant.tab.documents}"></h4>
<article
class="article"
th:data-url="@{${@pathRegistry.getPath(entityDto.targetEntity) + '/tab/documents'}}"
></article>
</div>
</th:block>
</html>
Template Structure:
-
Pending Decisions Section: Shows workflow status and available actions
-
Entity Details Tabs: Provides contextual information for decision-making
-
Dynamic Content Loading: Uses AJAX to load tab content dynamically
Template Features
Pending Decisions Fragment:
<th:block th:insert="~{/launchpad/fragment/pending-decisions :: body(${entityDto})}"/>
This inserts the standard pending decisions interface with entity-specific context.
Tabbed Content:
<article th:data-url="@{${@pathRegistry.getPath(entityDto.targetEntity) + '/tab/details'}}">
Dynamically loads tab content using the entity’s registered path.
10.6. Configuration
Register Your Extractor
Register your DecisionEntityExtractor as a Spring component:
@Component
public class ParticipantLaunchpadService implements DecisionEntityExtractor {
// Implementation...
}
The Launchpad system automatically discovers all DecisionEntityExtractor implementations and routes decisions accordingly.
10.7. Advanced Features
Custom Display Logic
Implement sophisticated display logic in your extractor:
@Override
public LaunchpadEntityDto getByHolderId(Long holderId) {
Participant participant = repository.findByPendingDecisionHolderId(holderId)
.orElseThrow(() -> new RuntimeException("No target entity with holder id " + holderId));
Application application = participant.getApplication();
String roles = participant.getRoles().stream()
.map(role -> EnumUtils.getLocalizedValue(role, LocaleContextHolder.getLocale()))
.collect(Collectors.joining(", "));
return new LaunchpadEntityDto(participant.getId(), application, application.getDisplayedName(),
participant, participant.getDisplayedName(), roles,
"/participant/launchpad-body");
}
private String buildContextualInfo(Participant participant) {
// Custom logic to build additional context
return participant.getStatus() + " | " +
participant.getPendingDocumentsCount() + " docs pending";
}
Multiple Entity Type Support
Create extractors for different entity types:
@Component
public class ApplicationLaunchpadService implements DecisionEntityExtractor {
@Override
public String getDecisionOwnerType() {
return Application.DECISION_OWNER_TYPE;
}
@Override
public LaunchpadEntityDto getByHolderId(Long holderId) {
// Implementation for application entities
}
}
@Component
public class CreditLaunchpadService implements DecisionEntityExtractor {
@Override
public String getDecisionOwnerType() {
return Credit.DECISION_OWNER_TYPE;
}
@Override
public LaunchpadEntityDto getByHolderId(Long holderId) {
// Implementation for credit entities
}
}
10.8. Best Practices
Entity Context Guidelines
Provide Rich Context: Include relevant parent/child relationships in the DTO
// Good: Include parent application for participant context
return new LaunchpadEntityDto(
participant.getId(),
application, // Parent context
application.getDisplayedName(),
participant, // Target entity
participant.getDisplayedName(),
roles,
templatePath
);
Meaningful Display Names: Use business-friendly names, not technical IDs
public String getDisplayedName() {
return String.format("%s (%s)", fullName, roles.iterator().next());
}
Localized Content: Support internationalization for role descriptions
String roles = participant.getRoles().stream()
.map(role -> EnumUtils.getLocalizedValue(role, LocaleContextHolder.getLocale()))
.collect(Collectors.joining(", "));
Performance Considerations
Efficient Queries: Use appropriate fetch strategies to avoid N+1 problems
@Query("SELECT p FROM Participant p " +
"JOIN FETCH p.application " +
"JOIN FETCH p.roles " +
"WHERE p.pendingDecisionHolder.id = :holderId")
Optional<Participant> findByPendingDecisionHolderId(@Param("holderId") Long holderId);
Caching: Consider caching frequently accessed context data
@Cacheable("participant-context")
public LaunchpadEntityDto getByHolderId(Long holderId) {
// Implementation...
}
10.9. Troubleshooting
Common Issues
ExtractorNotFoundException: No extractor found for decision owner type
Solution: Ensure your DecisionEntityExtractor is registered as @Component
and getDecisionOwnerType() returns the correct entity type constant.
Template Not Found: Launchpad template fragment missing
Solution: Verify template path in LaunchpadEntityDto matches actual file location.
Template must be in src/main/resources/templates/ directory.
Entity Not Found: PendingDecisionHolder ID doesn’t match entity
Solution: Check that findByPendingDecisionHolderId query is correct and
entity properly implements HasPendingDecisions interface.
The Launchpad system provides a powerful way to centralize decision management across different entity types, giving users a unified interface for workflow operations while maintaining entity-specific context and functionality.
11. Document Management
This section describes how to implement document management functionality for entities in the platform, including document type associations and UI integration.
11.1. Document System Overview
The document management system allows entities to have associated documents with configurable upload and requirement rules. The system consists of:
-
HasDocuments- Interface marking entities that can have documents -
DocumentTypeAssociation- Configuration for document types per entity -
EntityDocumentTabController- UI integration for document management tabs
11.2. Document Type Configuration
Document types are configured using DocumentTypeAssociation with a builder pattern that allows defining uploadable and required document types with conditional logic.
Document Type Associations
Required Document Configuration
Documents that must be uploaded based on entity conditions:
public static final EntityDocumentType OTHER = EntityDocumentType.OTHER;
public static final EntityDocumentType ID_SCAN = new EntityDocumentType("ID_SCAN");
private static final Predicate<Participant> PARTICIPANT_GUARANTOR =
participant -> participant.getRoles().contains(ParticipantRole.GUARANTOR);
private static final Predicate<Participant> PARTICIPANT_BORROWER =
participant -> participant.getRoles().contains(ParticipantRole.BORROWER);
@Bean
DocumentTypeAssociation<Participant> idScanDocumentTypeAssociations() {
return DocumentTypeAssociation.forEntityClass(Participant.class).required(ID_SCAN)
This configuration:
-
Makes
ID_SCANdocument required -
Only applies when participant status is
NEW -
Only applies to participants with
GUARANTORorBORROWERroles
11.3. UI Integration
Document Tab Implementation
To display document management interface, create a tab controller extending EntityDocumentTabController:
package com.timvero.example.admin.participant.tab;
import com.timvero.example.admin.participant.entity.Participant;
import com.timvero.web.common.tab.EntityDocumentTabController;
import org.springframework.core.annotation.Order;
import org.springframework.stereotype.Controller;
@Controller
@Order(1500)
public class ParticipantDocumentsTab extends EntityDocumentTabController<Participant> {
@Override
public boolean isVisible(Participant entity) {
return true;
}
}
Key features:
-
@Order(1500)- Controls tab display order in the UI -
isVisible()- Determines when the tab should be shown -
Automatic functionality - Upload, download, and delete operations are handled automatically
11.4. Builder Pattern Usage
The DocumentTypeAssociation uses a fluent builder pattern:
Available Methods
-
uploadable(EntityDocumentType)- Adds a document type that can be uploaded -
required(EntityDocumentType)- Adds a document type that must be uploaded -
predicate(Predicate<E>)- Adds conditional logic for when the association applies
Predicate Chaining
Multiple predicates can be combined:
new SignableDocumentType("APPLICATION_CONTRACT", ApplicationContractDocumentCategory.TYPE);
public static final EntityDocumentType OTHER = EntityDocumentType.OTHER;
public static final EntityDocumentType ID_SCAN = new EntityDocumentType("ID_SCAN");
Predicates are combined using AND logic - all conditions must be true.
11.5. Complete Implementation Example
To implement document management for an entity:
-
Entity implements HasDocuments:
@Entity
public class Participant implements HasDocuments {
// Entity implementation
}
-
Create document type configuration:
@Configuration
public class ParticipantDocumentTypesConfiguration {
@Bean
DocumentTypeAssociation<Participant> requiredDocuments() {
return DocumentTypeAssociation.forEntityClass(Participant.class)
.required(ID_SCAN)
.predicate(participant -> participant.getStatus() == ParticipantStatus.NEW)
.build();
}
@Bean
DocumentTypeAssociation<Participant> optionalDocuments() {
return DocumentTypeAssociation.forEntityClass(Participant.class)
.uploadable(OTHER)
.build();
}
}
-
Create document tab controller:
@Controller
@Order(1500)
public class ParticipantDocumentsTab extends EntityDocumentTabController<Participant> {
@Override
public boolean isVisible(Participant entity) {
return true; // Always show documents tab
}
}
This provides a complete document management system with conditional requirements and integrated UI.
12. Document Templates
This section covers creating and customizing document templates for contracts, reports, and other generated documents in the platform.
12.1. Template System Overview
The document template system enables automatic generation of formatted documents using entity data. The system consists of:
-
DocumentCategory- Defines document types and data models -
DocumentTemplate- Stores template content (HTML/TXT) -
DocumentModel- Provides data structure for templates -
Template Engine - Processes templates with entity data
12.2. Creating Document Categories
Document categories define what data is available to templates and when documents can be generated.
Basic Document Category
import java.util.UUID;
import org.springframework.stereotype.Component;
@Component
public class ApplicationContractDocumentCategory
extends DocumentCategory<UUID, Participant, ContractDocumentModel> {
public static final DocumentType TYPE = new DocumentType("APPLICATION_CONTRACT");
@Override
public DocumentType getType() {
return TYPE;
}
@Override
protected ContractDocumentModel getModel(Participant participant) {
PaymentSchedule paymentSchedule = participant.getApplication().getPaymentSchedule();
return new ContractDocumentModel(participant.getApplication(), paymentSchedule);
}
Key components:
-
DocumentType- Unique identifier for the document type -
getModel()- Provides data structure for template processing -
isSuitableTestEntity()- Determines when document generation is available
Document Model Creation
The document model exposes entity data to templates:
@Override
protected boolean isSuitableTestEntity(Participant participant) {
return participant.getStatus() == ParticipantStatus.APPROVED
&& participant.getApplication().getPaymentSchedule() != null;
}
public static class ContractDocumentModel extends DocumentModel {
private Application application;
private PaymentSchedule paymentSchedule;
public ContractDocumentModel() {
}
public ContractDocumentModel(Application application, PaymentSchedule paymentSchedule) {
this.application = application;
this.paymentSchedule = paymentSchedule;
}
public Application getApplication() {
return application;
}
public PaymentSchedule getPaymentSchedule() {
return paymentSchedule;
}
Model features:
-
Public getters - Expose data to template engine
-
Computed properties - Derive values from entity data (e.g.,
getFirstPayment()) -
Nested objects - Include related entities for complex templates
12.3. Template Content Creation
Templates are created using the WYSIWYG editor in the admin interface, which provides rich text formatting and HTML support.
Using the WYSIWYG Editor
The admin interface provides a visual editor for creating document templates:
-
Rich Text Formatting - Bold, italic, fonts, colors, alignment
-
HTML Support - Full HTML markup for advanced formatting
-
Variable Insertion - Insert dynamic variables using Pebble syntax
-
Preview Mode - See how templates render with sample data
Pebble Template Processing
Templates use Pebble template engine for dynamic content processing. Pebble provides:
-
Variable substitution -
{{ variable.property }} -
Conditional logic -
{% if condition %}…{% endif %} -
Loops -
{% for item in items %}…{% endfor %} -
Filters -
{{ amount | currency }},{{ date | date('yyyy-MM-dd') }}
Basic Template Example
<h1>Loan Agreement</h1>
<p>Borrower: {{ application.client.individualInfo.firstName }} {{ application.client.individualInfo.lastName }}</p>
<p>Loan Amount: <strong>{{ application.requestedAmount }}</strong></p>
<h3>Payment Schedule</h3>
<table border="1">
<tr><th>Payment Date</th><th>Amount</th></tr>
{% for payment in paymentSchedule.payments.values %}
<tr>
<td>{{ payment.dueDate | date('yyyy-MM-dd') }}</td>
<td>{{ payment.amount }}</td>
</tr>
{% endfor %}
</table>
12.4. Template Variables and Functions
Available Variables
Templates have access to:
-
Model properties - All public getters from your DocumentModel
-
Built-in filters - Date formatting, number formatting, string manipulation
-
Global variables - Current date/time, system settings
Common Pebble Filters
Date Formatting:
{{ payment.dueDate | date('MMMM dd, yyyy') }} // March 15, 2024
{{ payment.dueDate | date('yyyy-MM-dd') }} // 2024-03-15
Number Formatting:
{{ amount | currency }} // $1,234.56
{{ interestRate | numberformat('#,##0.00%') }} // 5.50%
{{ amount | numberformat('#,##0.00') }} // 1,234.56
String Formatting:
{{ client.name | upper }} // JOHN DOE
{{ client.name | lower }} // john doe
{{ client.name | title }} // John Doe
Conditional Content:
{% if application.status == 'APPROVED' %}
This loan has been approved.
{% else %}
This loan is pending review.
{% endif %}
Loops:
{% for participant in participants %}
{{ participant.role }}: {{ participant.client.displayName }}
{% endfor %}
12.5. Template Management
12.6. Advanced Features
Document Queries with EntityDocumentFinder
The EntityDocumentFinder service provides powerful document querying capabilities for conditional template logic and document management.
Document Type Distinction
The system has two types of documents:
-
EntityDocumentType- Regular uploaded documents (PDFs, images, etc.) -
SignableDocumentType- Generated documents that require signatures
Methods work with different types:
-
Document existence:
EntityDocumentType -
Signature operations:
SignableDocumentTypeonly
Checking Document Existence
Use EntityDocumentFinder to check if documents exist before generation:
@Service
public class ContractGenerationService {
@Autowired
private EntityDocumentFinder documentFinder;
// Document type constants
private static final EntityDocumentType ID_SCAN = new EntityDocumentType("ID_SCAN");
private static final EntityDocumentType INCOME_PROOF = new EntityDocumentType("INCOME_PROOF");
private static final SignableDocumentType CONTRACT_TYPE = new SignableDocumentType("CONTRACT");
public boolean canGenerateContract(Participant participant) {
// Check if required documents are present (EntityDocumentType)
boolean hasIdScan = documentFinder.isPresent(participant, ID_SCAN);
boolean hasIncomeProof = documentFinder.isPresent(participant, INCOME_PROOF);
// Check if contract already exists and is signed (SignableDocumentType)
boolean contractSigned = documentFinder.isLatestSigned(participant, CONTRACT_TYPE);
return hasIdScan && hasIncomeProof && !contractSigned;
}
}
Document Status in Templates
Access document information in your DocumentModel:
public class ContractDocumentModel extends DocumentModel {
private Participant participant;
private EntityDocumentFinder documentFinder;
// Document type constants
private static final SignableDocumentType CONTRACT_TYPE = new SignableDocumentType("CONTRACT");
private static final SignableDocumentType PREVIOUS_CONTRACT_TYPE = new SignableDocumentType("PREVIOUS_CONTRACT");
public ContractDocumentModel(Participant participant, EntityDocumentFinder finder) {
this.participant = participant;
this.documentFinder = finder;
}
public boolean hasSignedPreviousContract() {
return documentFinder.isLatestSigned(participant, PREVIOUS_CONTRACT_TYPE);
}
public String getLastContractDate() {
return documentFinder.latest(participant, CONTRACT_TYPE)
.map(doc -> doc.getCreatedAt().format(DateTimeFormatter.ofPattern("yyyy-MM-dd")))
.orElse("No previous contract");
}
}
Use in templates:
{% if hasSignedPreviousContract %}
<p>This replaces your previous contract dated {{ lastContractDate }}.</p>
{% else %}
<p>This is your first contract with us.</p>
{% endif %}
Available Query Methods
Document Existence (EntityDocumentType):
-
isPresent(owner, EntityDocumentType)- Check if document exists -
isMissing(owner, EntityDocumentType)- Check if document is missing
Document Retrieval:
-
getDocuments(owner)- Get all active documents -
latest(owner, EntityDocumentType)- Get most recent document of type -
latest(owner, SignableDocumentType)- Get most recent signable document
Signature Status (SignableDocumentType only):
-
isLatestSigned(owner, SignableDocumentType)- Check if latest document is signed -
signatureOfLatest(owner, SignableDocumentType)- Get signature details
Multiple Document Templates
A single DocumentCategory can have multiple templates created in the admin interface. Generate documents using specific template IDs:
Single Document Category
@Component
public class ContractDocumentCategory extends DocumentCategory<UUID, Application, ContractDocumentModel> {
public static final DocumentType CONTRACT = new DocumentType("CONTRACT");
@Override
public DocumentType getType() {
return CONTRACT;
}
@Override
protected ContractDocumentModel getModel(Application application) {
return new ContractDocumentModel(application, application.getPaymentSchedule());
}
@Override
protected boolean isSuitableTestEntity(Application application) {
return application.getStatus() == ApplicationStatus.APPROVED;
}
}
Document Model with Conditional Logic
The DocumentModel can expose properties for template conditional logic:
public class ContractDocumentModel extends DocumentModel {
private Application application;
private PaymentSchedule schedule;
private LocalDate generationDate;
public ContractDocumentModel(Application application, PaymentSchedule schedule) {
this.application = application;
this.schedule = schedule;
this.generationDate = LocalDate.now();
}
public Application getApplication() {
return application;
}
public PaymentSchedule getSchedule() {
return schedule;
}
// Template conditional properties
public boolean isSecuredLoan() {
return application.isSecuredLoan();
}
public boolean isHighValueLoan() {
return application.getRequestedAmount().isGreaterThan(MonetaryAmount.of(100000, "USD"));
}
public String getContractType() {
if (isSecuredLoan()) {
return "Secured Loan Agreement";
}
return "Standard Loan Agreement";
}
// Secured loan specific data (null if not secured)
public String getCollateralDescription() {
return application.getCollateral() != null ?
application.getCollateral().getDescription() : null;
}
}
Multiple Templates in Admin Interface
Create different templates for the same DocumentCategory:
-
Standard Contract Template:
-
Name: "Standard Loan Contract"
-
Document Type: CONTRACT
-
Content: Basic loan terms without collateral sections
-
-
Secured Contract Template:
-
Name: "Secured Loan Contract"
-
Document Type: CONTRACT
-
Content: Includes collateral information using conditional logic
-
-
High-Value Contract Template:
-
Name: "High-Value Loan Contract"
-
Document Type: CONTRACT
-
Content: Additional compliance sections for large loans
-
Template Conditional Content
Templates use the same DocumentModel but show different content:
Standard Contract Template:
<h1>{{ contractType }}</h1>
<p>Loan Amount: {{ application.requestedAmount }}</p>
<p>Interest Rate: {{ application.condition.interestRate }}%</p>
{% if not isSecuredLoan %}
<h3>Unsecured Loan Terms</h3>
<p>This loan is not secured by collateral...</p>
{% endif %}
Secured Contract Template:
<h1>{{ contractType }}</h1>
<p>Loan Amount: {{ application.requestedAmount }}</p>
<p>Interest Rate: {{ application.condition.interestRate }}%</p>
{% if isSecuredLoan %}
<h3>Collateral Information</h3>
<p><strong>Description:</strong> {{ collateralDescription }}</p>
<p><strong>Value:</strong> {{ application.collateral.estimatedValue }}</p>
{% endif %}
Generating with Specific Templates
Generate documents using template IDs:
// Generate using specific template
SignableDocument document = documentService.generate(
application,
CONTRACT,
standardContractTemplateId // Specify which template to use
);
// Or for secured loans
SignableDocument securedDocument = documentService.generate(
application,
CONTRACT,
securedContractTemplateId // Different template, same DocumentCategory
);
This approach allows:
-
Single DocumentCategory - One category handles all contract variations
-
Multiple templates - Different layouts and content for same data
-
Template selection - Choose appropriate template based on business logic
-
Shared data model - Same DocumentModel serves all template variations
Custom Template Functions
Extend template capabilities with custom functions in your DocumentModel:
public class ContractDocumentModel extends DocumentModel {
// Format complex data for display
public String getFormattedLoanTerm() {
int months = application.getCondition().getTermInMonths();
return months == 12 ? "1 year" : months + " months";
}
// Business logic for conditional content
public boolean isHighRiskLoan() {
return application.getCondition().getInterestRate() > 15.0;
}
// Computed properties
public MonetaryAmount getMonthlyPayment() {
return schedule.getPayments().values().stream()
.findFirst()
.map(PaymentSegment::getAmount)
.orElse(MonetaryAmount.ZERO);
}
// Formatting helpers
public String getClientFullName() {
IndividualInfo info = application.getClient().getIndividualInfo();
return info.getFirstName() + " " + info.getLastName();
}
// Status checks
public boolean requiresAdditionalDocumentation() {
return isHighValueLoan() || isHighRiskLoan();
}
}
Use in templates:
<h1>{{ contractType }} - {{ formattedLoanTerm }}</h1>
<p>Borrower: {{ clientFullName }}</p>
<p>Monthly Payment: {{ monthlyPayment | currency }}</p>
{% if highRiskLoan %}
<div class="warning">
<strong>High Risk Loan</strong> - Additional terms apply
</div>
{% endif %}
{% if requiresAdditionalDocumentation %}
<p><em>Additional documentation may be required.</em></p>
{% endif %}
12.7. Integration with Signature Workflow
Signable Documents
For documents requiring signatures:
// Document generation creates SignableDocument
SignableDocument document = documentService.generate(
participant,
APPLICATION_CONTRACT,
templateId
);
// Document flows through signature process
// PENDING_SIGNATURE -> SIGNED -> contract complete
Signature Status Lifecycle
Documents progress through signature states:
-
GENERATION_FAILED- Document generation failed -
PENDING_SIGNATURE- Waiting for signature -
SIGNED- Successfully signed -
REFUSE- Signature refused -
REVOKE- Signature revoked
DocumentSignature Extensions
The system supports different signature types through inheritance. Create custom signature implementations for specific workflows:
Physical Document Signature
For in-person or printed document signing:
@Entity
@Table(name = "physical_document_signature")
@DiscriminatorValue("PHYSICAL")
public class PhysicalDocumentSignature extends DocumentSignature {
}
Custom Signature Types
Create specialized signature classes for different signing methods:
@Entity
@Table(name = "electronic_document_signature")
@DiscriminatorValue("ELECTRONIC")
public class ElectronicDocumentSignature extends DocumentSignature {
@Column(name = "ip_address")
private String ipAddress;
@Column(name = "user_agent")
private String userAgent;
@Column(name = "signature_timestamp")
private Instant signatureTimestamp;
@Column(name = "verification_code")
private String verificationCode;
// Getters and setters
public String getIpAddress() { return ipAddress; }
public void setIpAddress(String ipAddress) { this.ipAddress = ipAddress; }
public String getUserAgent() { return userAgent; }
public void setUserAgent(String userAgent) { this.userAgent = userAgent; }
public Instant getSignatureTimestamp() { return signatureTimestamp; }
public void setSignatureTimestamp(Instant timestamp) { this.signatureTimestamp = timestamp; }
public String getVerificationCode() { return verificationCode; }
public void setVerificationCode(String code) { this.verificationCode = code; }
}
Signature Actions
Create action controllers to handle signature workflows:
@RequestMapping("/sign")
@Controller
@Order(1000)
public class SignDocumentAction extends SimpleActionController<UUID, SignableDocument> {
@Autowired
private ParticipantRepository participantRepository;
@Autowired
private EntityDocumentService entityDocumentService;
@Override
protected EntityAction<SignableDocument, Object> action() {
return when(d -> isRequiredDocAdded(d)
&& d.getStatus().in(SignatureStatus.PENDING_SIGNATURE)
&& d.getDocument() != null)
.then((document, f, u) -> {
Participant participant = participantRepository.getReferenceById(document.getOwnerId());
// Create appropriate signature type
PhysicalDocumentSignature signature = new PhysicalDocumentSignature();
signature.setSigner(participant.getDisplayedName());
signature.setEmail(participant.getClient().getContactInfo().getEmail());
signature.setPhone(participant.getClient().getContactInfo().getPhone());
document.setSignature(signature);
document.setStatus(SignatureStatus.SIGNED);
});
}
private boolean isRequiredDocAdded(SignableDocument d) {
Participant participant = participantRepository.getReferenceById(d.getOwnerId());
return entityDocumentService.requiredDocumentsAdded(participant);
}
}
Signature Integration
Templates can include signature placeholders and access signature information:
<div class="signature-section">
<p>Borrower Signature:</p>
<div class="signature-line">_________________________</div>
<p>Date: {{ generationDate }}</p>
{% if capturedSignature %}
<p><strong>Signed by:</strong> {{ signerName }}</p>
<p><strong>Signature Type:</strong> {{ capturedSignature.class.simpleName }}</p>
{% endif %}
</div>
Static Date Handling: Capture the generation date in the DocumentModel constructor, not in a getter:
public class ContractDocumentModel extends DocumentModel {
private LocalDate generationDate;
public ContractDocumentModel(Application application) {
this.generationDate = LocalDate.now(); // Captured at generation time
}
public String getGenerationDate() {
return generationDate.format(DateTimeFormatter.ofPattern("yyyy-MM-dd"));
}
}
12.8. Complete Implementation Example
To implement a new document template:
-
Create DocumentCategory:
@Component
public class LoanSummaryDocumentCategory extends DocumentCategory<UUID, Application, LoanSummaryModel> {
public static final DocumentType LOAN_SUMMARY = new DocumentType("LOAN_SUMMARY");
@Autowired
private EntityDocumentFinder documentFinder;
@Override
public DocumentType getType() {
return LOAN_SUMMARY;
}
@Override
protected LoanSummaryModel getModel(Application application) {
// Capture dynamic data at generation time
int docCount = documentFinder.getDocuments(application).size();
return new LoanSummaryModel(application, application.getPaymentSchedule(), docCount);
}
@Override
protected boolean isSuitableTestEntity(Application application) {
// Note: Using EntityDocumentType for document existence check
EntityDocumentType contractEntityType = new EntityDocumentType("CONTRACT");
return application.getStatus() == ApplicationStatus.SERVICING
&& documentFinder.isPresent(application, contractEntityType);
}
}
-
Create Document Model (static data only):
public class LoanSummaryModel extends DocumentModel {
private Application application;
private PaymentSchedule schedule;
private LocalDate generationDate;
private int documentCount; // Captured at generation time
public LoanSummaryModel(Application application, PaymentSchedule schedule, int docCount) {
this.application = application;
this.schedule = schedule;
this.generationDate = LocalDate.now(); // Fixed at generation
this.documentCount = docCount; // Static snapshot
}
public MonetaryAmount getTotalInterest() {
return schedule.getPayments().values().stream()
.map(PaymentSegment::getInterestAmount)
.reduce(MonetaryAmount.ZERO, MonetaryAmount::add);
}
public String getGenerationDate() {
return generationDate.format(DateTimeFormatter.ofPattern("yyyy-MM-dd"));
}
public int getDocumentCount() {
return documentCount; // Static value
}
}
-
Create HTML Template in admin interface using model properties
-
Generate Documents using
DocumentService.generate()
This provides a complete document template system with rich formatting, dynamic content, and signature integration.
13. Notifications
This section covers creating and managing automated notifications for loan applications and credit accounts. The notification system provides multi-channel communication (EMAIL, SMS, VOICE, LETTER) with template-based content.
13.1. Notification System Overview
The notification system consists of:
-
NotificationEvent- Defines when notifications are triggered -
NotificationModel- Provides data to notification templates -
NotificationTemplate- Database-stored templates with dynamic content -
Notification Gateways - Handle delivery via EMAIL, SMS, VOICE, LETTER
13.2. Creating Notification Events
Notification events define when and how notifications are sent to users.
Basic Notification Event
@Component
public class LoanApprovedEvent extends NotificationEvent<Application, NotificationModel> {
public static final NotificationEventType TYPE = new NotificationEventType("LOAN_APPROVED");
@Override
public NotificationEventType getType() {
return TYPE;
}
public boolean notify(UUID applicationId) {
return super.notify(applicationId, new NotificationModel());
}
@Override
protected boolean isSuitableTestEntity(Application application) {
return application.getStatus() == ApplicationStatus.SERVICING;
Key components:
-
NotificationEventType- Unique identifier for the event type -
getType()- Returns the event type for template matching -
notify()- Triggers notification with entity data -
isSuitableTestEntity()- Determines when event can be tested
Event with Custom Data
public class LoanDeclinedEvent extends NotificationEvent<Application, LoanDeclineTemplateModel> {
public static final NotificationEventType TYPE = new NotificationEventType("LOAN_DECLINED");
@Override
public NotificationEventType getType() {
return TYPE;
}
public boolean notify(UUID applicationId, DeclineReason declineReason, boolean reapplicationEligible) {
LoanDeclineTemplateModel model = new LoanDeclineTemplateModel();
model.setDeclineReason(declineReason);
model.setReapplicationEligible(reapplicationEligible);
return super.notify(applicationId, model);
}
@Override
protected boolean isSuitableTestEntity(Application application) {
Custom template models provide specific data for templates:
private DeclineReason declineReason;
private boolean reapplicationEligible;
public DeclineReason getDeclineReason() {
return declineReason;
}
public void setDeclineReason(DeclineReason declineReason) {
this.declineReason = declineReason;
}
public boolean isReapplicationEligible() {
return reapplicationEligible;
}
public void setReapplicationEligible(boolean reapplicationEligible) {
this.reapplicationEligible = reapplicationEligible;
}
public Integer getReapplicationWaitDays() {
return reapplicationEligible ? 30 : 180;
Event with Parameters
Events can use EventParameters for template-configurable logic:
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;
@Component
public class PaymentReminderEvent extends NotificationEvent<ExampleCredit, PaymentReminderTemplateModel>
implements EntityEventListener<CreditCalculationDateChangeEvent> {
public static final NotificationEventType TYPE = new NotificationEventType("PAYMENT_REMINDER");
private static final String DAYS_COUNT = "daysCount";
private static final EventParameter DAYS_COUNT_PARAM = EventParameter.integerField("daysCount");
@Autowired
private PastDueOperationService pastDueService;
EventParameters allow admin users to configure notification behavior:
-
daysCount- Configurable number of days for payment reminders -
Used in
entityMatchesTemplate()to control when notifications are sent
13.3. Creating Notification Templates
Templates are created and managed through the admin interface, similar to document templates.
Creating New Template
-
Template Details:
-
Name: Descriptive name (e.g., "Loan Approval - Email")
-
Event Type: Select from dropdown (e.g., "LOAN_APPROVED")
-
Gateway Type: EMAIL, SMS, VOICE, or LETTER
-
Recipient Type: BORROWER, AGENT, etc.
-
Active: Check to enable template
-
-
Template Content: Use WYSIWYG editor with Pebble template syntax
Template Content Examples
Email Template (Loan Approval) - Minimal
Subject: Loan Approved - {{ application.id | slice(0, 8) | upper }}
Body:
<h2>Congratulations {{ application.client.individualInfo.firstName }}!</h2>
<p>Your {{ application.requestedAmount | currency }} loan has been approved at {{ application.condition.interestRate | numberformat('#,##0.00%') }}.</p>
<p>Documents will be ready within 2 business days.</p>
SMS Template (Payment Reminder)
{% if reminderType == 'UPCOMING' %}
Payment due {{ expectedPaymentDate | date('MM/dd') }}.
{% else %}
Payment overdue by {{ daysCount | abs }} days.
{% endif %}
Account: {{ credit.id | slice(0, 8) | upper }}
Pay online or call (555) 123-4567
Email Template (Loan Decline)
Subject: Loan Application Update
Body:
<h2>Dear {{ application.client.individualInfo.firstName }},</h2>
{% if declineReason %}
<p><strong>Status:</strong> {{ declineReason.description }}</p>
{% else %}
<p>Your application did not meet our current lending criteria.</p>
{% endif %}
<p>You may reapply after {{ reapplicationWaitDays }} days.</p>
<p>Questions? Call (555) 123-4567.</p>
13.4. Template Variables and Functions
Available Variables
Templates have access to:
-
Entity data - All properties from the associated entity (Application, ExampleCredit, etc.)
-
Model data - Custom data from NotificationModel subclasses
-
Built-in filters - Date formatting, number formatting, string manipulation
Common Pebble Filters
Date Formatting:
{{ application.createdAt | date('MMMM dd, yyyy') }} // March 15, 2024
{{ expectedPaymentDate | date('yyyy-MM-dd') }} // 2024-03-15
Number Formatting:
{{ application.requestedAmount | currency }} // $50,000.00
{{ application.condition.interestRate | numberformat('#,##0.00%') }} // 5.50%
String Operations:
{{ application.id | slice(0, 8) | upper }} // A1B2C3D4
{{ application.client.individualInfo.firstName | title }} // John
Conditional Content:
{% if application.status == 'SERVICING' %}
Your loan is now active.
{% else %}
Your application is being processed.
{% endif %}
Loops (for collections):
{% for document in application.documents %}
Document: {{ document.name }}
{% endfor %}
13.5. Event Parameters vs Custom Models
Use EventParameters when: Admin users need to configure when notifications send (timing, thresholds, conditions)
Use Custom Models when: Templates need additional data not available on the base entity
EventParameter Example
Payment reminders need admin-configurable timing:
public static final EventParameter DAYS_COUNT = EventParameter.integerField("daysCount");
public PaymentReminderEvent() {
super(DAYS_COUNT);
}
@Override
protected boolean entityMatchesTemplate(ExampleCredit credit, PaymentReminderTemplateModel model,
Map<String, Object> templateParams) {
Integer daysCount = (Integer) templateParams.get("daysCount");
model.setDaysCount(daysCount);
if (daysCount < 0) {
// Overdue reminder: check if credit is exactly |daysCount| days past due
return pastDueService.daysPastDue(credit) == -daysCount;
} else {
// Upcoming reminder: check if payment is exactly daysCount days away
return getNextPaymentDate(credit).equals(today.plusDays(daysCount));
}
}
13.6. Automatic Notifications
Entity Event Listeners
Events can implement EntityEventListener to automatically trigger on system events:
@Component
public class PaymentReminderEvent extends NotificationEvent<ExampleCredit, PaymentReminderTemplateModel>
implements EntityEventListener<CreditCalculationDateChangeEvent> {
public PaymentReminderEvent() {
super(EventParameter.integerField("daysCount"));
}
@Override
public void handle(CreditCalculationDateChangeEvent event) {
notify(event.getCreditId(), new PaymentReminderTemplateModel());
}
@Override
protected boolean entityMatchesTemplate(ExampleCredit credit, PaymentReminderTemplateModel model,
Map<String, Object> templateParams) {
Integer daysCount = (Integer) templateParams.get("daysCount");
model.setDaysCount(daysCount);
if (daysCount < 0) {
// Overdue: check if exactly |daysCount| days past due
return pastDueService.daysPastDue(credit) == -daysCount;
} else {
// Upcoming: check if payment is exactly daysCount days away
return ChronoUnit.DAYS.between(today, nextPaymentDate) == daysCount;
}
}
}
This automatically triggers payment reminders when credit calculation dates change, with templates configured for specific day thresholds.
EntityChecker Integration
Notifications can be triggered automatically using EntityCheckers:
@Component
public class ApplicationStatusNotificationChecker extends EntityChecker<UUID, Application> {
@Autowired
private LoanApprovedEvent loanApprovedEvent;
@Autowired
private LoanDeclinedEvent loanDeclinedEvent;
@Override
protected void registerListeners() {
register(Application.class, EntityChangeEvent.Operation.UPDATE);
}
@Override
protected boolean isAvailable(Application application) {
return application.getStatus().in(ApplicationStatus.SERVICING, ApplicationStatus.DECLINE);
}
@Override
protected void perform(Application application) {
switch (application.getStatus()) {
case SERVICING -> loanApprovedEvent.notify(application.getId());
case DECLINE -> loanDeclinedEvent.notify(application.getId(),
application.getDeclineReason(), true);
}
}
}
13.7. Notification Gateways
Gateways handle actual delivery of notifications via EMAIL, SMS, VOICE, or LETTER channels.
Gateway Interface
public interface NotificationGateway {
NotificationGatewayType getType();
String send(Notification notification) throws Exception;
Map<String, NotificationStatus> check(Collection<String> messageIds) throws Exception;
}
Custom Implementation
Extend abstract gateways for specific providers:
@Component
public class TwilioSmsGateway extends AbstractSmsGateway {
@Override
protected String sendSms(String phone, String message) throws Exception {
// Integrate with Twilio API
Message twilioMessage = Message.creator(
new PhoneNumber(phone),
new PhoneNumber(fromPhoneNumber),
message
).create();
return twilioMessage.getSid();
}
@Override
public Map<String, NotificationStatus> check(Collection<String> messageIds) throws Exception {
// Check delivery status from Twilio
return messageIds.stream().collect(Collectors.toMap(
id -> id,
id -> getDeliveryStatus(id)
));
}
}
13.8. Internationalization (i18n)
The notification system integrates with Spring’s message system for localized labels and descriptions.
Event Type Labels
Event types are displayed in the admin interface using message keys:
# messages.properties
notification.event.type.APP_STATUS_CHANGE=Application Status Change
notification.event.type.PAYMENT_REMINDER=Payment Reminder
notification.event.type.LOAN_APPROVED=Loan Approved
notification.event.type.LOAN_DECLINED=Loan Declined
The pattern is: notification.event.type.{EVENT_TYPE_NAME}=Display Name
EventParameter Labels
EventParameter fields are labeled using their parameter names:
# messages.properties
notification.event.APP_STATUS_CHANGE.status=Status
notification.event.PAYMENT_REMINDER.daysCount=Days Count
notification.event.LOAN_DECLINED.declineReason=Decline Reason
The pattern is: notification.event.{EVENT_TYPE_NAME}.{PARAMETER_NAME}=Field Label
13.9. Testing Notifications
Template Testing
Use the admin interface to test templates:
-
Go to Notification Templates
-
Select a template
-
Click "Test Template"
-
Choose a suitable entity from the dropdown (filtered by
isSuitableTestEntity()) -
Preview the rendered template
Programmatic Testing
@Test
public void testLoanApprovalNotification() {
// Create test application
Application application = createTestApplication();
application.setStatus(ApplicationStatus.SERVICING);
// Test notification
boolean sent = loanApprovedEvent.notify(application.getId());
assertTrue(sent);
}
13.10. Integration Examples
Scheduled Payment Reminders
@Component
public class PaymentReminderScheduler {
@Autowired
private PaymentReminderEvent paymentReminderEvent;
@Scheduled(cron = "0 0 9 * * ?") // Daily at 9 AM
public void sendPaymentReminders() {
// PaymentReminderEvent handles the business logic
// Templates configured with different daysCount values will
// automatically send appropriate reminders
List<ExampleCredit> activeCredits = creditRepository.findActiveCredits();
for (ExampleCredit credit : activeCredits) {
paymentReminderEvent.notify(credit.getId());
}
}
}
Workflow Integration
@Component
public class WorkflowCompletionListener implements EntityEventListener<ProcessExecutionFinishedEvent> {
@Autowired
private LoanApprovedEvent loanApprovedEvent;
@Override
public void handle(ProcessExecutionFinishedEvent event) {
if ("LOAN_APPROVAL_PROCESS".equals(event.getProcessDefinitionKey())) {
if ("APPROVED".equals(event.getResult())) {
loanApprovedEvent.notify(event.getEntityId());
}
}
}
}
13.11. Common Patterns
Event Design Decisions
// ✅ Good: Simple event, no custom data needed
public class LoanApprovedEvent extends NotificationEvent<Application, NotificationModel> {
public boolean notify(UUID applicationId) {
return super.notify(applicationId, new NotificationModel());
}
}
// ✅ Good: Custom data for template logic
public class LoanDeclinedEvent extends NotificationEvent<Application, LoanDeclineTemplateModel> {
public boolean notify(UUID applicationId, DeclineReason reason, boolean eligible) {
LoanDeclineTemplateModel model = new LoanDeclineTemplateModel();
model.setDeclineReason(reason);
model.setReapplicationEligible(eligible);
return super.notify(applicationId, model);
}
}
// ❌ Avoid: EventParameter for static data
public class BadEvent extends NotificationEvent<Application, NotificationModel> {
private static final EventParameter COMPANY_NAME = EventParameter.stringField("companyName");
// Use application properties or constants instead
}
Template Patterns
<!-- ✅ Good: Fallback for missing data -->
{{ application.client.individualInfo.firstName | default("Valued Customer") }}
<!-- ✅ Good: Safe navigation with conditionals -->
{% if application.condition %}
Rate: {{ application.condition.interestRate | numberformat('#,##0.00%') }}
{% endif %}
<!-- ❌ Avoid: Complex calculations in templates -->
<!-- Move business logic to NotificationModel methods -->
Integration Patterns
Automatic notifications: Use EntityChecker for state-driven events Scheduled reminders: Use EventParameters for admin-configurable timing Workflow notifications: Listen to process completion events
Performance Notes
-
Batch scheduled notifications - don’t send individually in loops
-
Use isSuitableTestEntity() to filter test data accurately
-
Monitor gateway delivery rates - EMAIL typically higher success than SMS
This notification system provides flexible, template-based communication that integrates seamlessly with your loan management workflow.
14. Entity Checkers — setup & usage
This section describes how to set up and manage Entity Checkers for automated business rule validation and state management in the application.
14.1. Checker System Architecture
The platform uses Entity Checkers to implement event-driven business logic that automatically responds to entity changes. Checkers are reactive components that listen to DB changes and execute rules when conditions are met.
What are Entity Checkers?
-
Monitor Entity Changes — detect create/update/delete of entities
-
Apply Business Rules — execute predefined logic on matched conditions
-
Maintain Data Consistency — keep related entities in sync
-
Automate Workflows — trigger next steps without manual actions
Checker Class Hierarchy
-
Base Checker Classes — framework abstract classes (e.g.
EntityChecker) provide infrastructure -
Custom Entity Checkers — app-specific implementations e.g.
BorrowerStartTreeChecker— manages borrower workflow start
Each checker has three core parts:
-
Listener Registration (
registerListeners) — what changes to monitor -
Availability Check (
isAvailable) — when the checker should run -
Business Logic (
perform) — what to do if available
14.2. Listener Registration
registerListeners defines which changes trigger the checker, configured via CheckerListenerRegistry.
CheckerListenerRegistry
CheckerListenerRegistry<E> configures listeners for entity changes where E is the target entity type (e.g., Application, Participant).
| Method | Parameters | Description |
|---|---|---|
|
|
Create a listener for changes on a different type. See Using |
|
(none) |
Monitor the same entity type as the checker (shorthand). |
|
|
Filter to field updates only. |
|
(none) |
Filter to insert events only. |
|
|
Add custom filter predicates; can be chained. |
Using entityChange method
Problem: the checker monitors the same type it operates on. Solution:
registry.entityChange().updated("status");
-
Trigger entity == target entity
-
Direct field monitoring
-
Simple relationship context
Problem: monitor a related entity and map to the target.
registry.entityChange(Participant.class, Participant::getApplication)
.updated("status");
-
Direct JPA relationships (1:1, N:1)
-
Same transaction context
-
Clear getter mapping
Problem: mapping requires repository lookup.
registry.entityChange(SignableDocument.class,
d -> participantRepository.getReferenceById(d.getOwnerId()))
.updated("status");
|
Performance: repository lookups add DB queries — use only if no direct association is available. |
14.3. Checker Implementation Examples
BorrowerStartTreeChecker
Manages borrower workflow initiation when participants complete required documentation.
| Component | Description |
|---|---|
Target Entity |
|
Purpose |
Auto-start decision tree when borrower completes requirements |
Triggers |
Document signatures and required uploads |
Business Logic |
Set status to |
Listener 1: Application Form Signature Monitor
Purpose: Track completion of form signatures
Trigger: SignableDocument.status → SIGNED
Target: Map document → owning participant
registry.entityChange(SignableDocument.class, d -> participantRepository.getReferenceById(d.getOwnerId()))
.updated(SignableDocument_.STATUS)
.and(d -> d.getStatus() == SignatureStatus.SIGNED && d.getDocumentType() == ParticipantDocumentTypesConfiguration.APPLICATION_FORM);
Listener 2: Required Document Upload Monitor
Purpose: Detect required document uploads
Trigger: New EntityDocument insertions of required types
Target: Map upload → owning participant
registry.entityChange(EntityDocument.class, d -> participantRepository.getReferenceById(d.getOwnerId()))
.inserted().and(d -> {
Participant participant = participantRepository.getReferenceById(d.getOwnerId());
return documentService.getRequiredDocumentTypes(participant).contains(d.getDocumentType());
});
Availability Check
@Override
protected boolean isAvailable(Participant participant) {
return needSignature(participant) && hasSignature(participant);
}
-
Participant is a
BORROWER -
Status is
NEW -
Application form is signed
-
All required documents uploaded
Business Logic
@Override
protected void perform(Participant participant) {
participant.setStatus(ParticipantStatus.IN_PROCESS);
decisionProcessStarter.start(PARTICIPANT_TREE, participant.getId());
}
-
Status Update —
NEW→IN_PROCESS -
Process Initiation — start automated decision tree
-
Transaction Safety — single transaction
14.4. Best Practices
Design & Registration
-
Listen narrowly — subscribe only to fields/states that matter
-
Prefer direct mapping — use entity associations before repo lookups
-
Add predicates — guard with
and(…)to avoid noisy triggers -
Idempotent perform() — safe on duplicate events/retries
14.5. Anti-Patterns
-
{bad} Using broad listeners — catching all updates without field filters
-
{bad} Heavy logic in
isAvailable()— long DB scans/joins -
{bad} Side-effects in availability — mutate state during checks
-
{bad} Repository mapping by default — use only when no direct relation
-
{bad} Non-idempotent
perform()— repeated activations break state -
{bad} Silent failures — no logs/metrics on skip/error
14.6. Production Checklist
-
{todo} Narrow listeners to specific fields/states
-
{todo} Add
and(…)predicates to reduce noise -
{todo} Make
perform()idempotent and transactional -
{todo} Ensure
isAvailable()is read-only and fast -
{todo} Add logs/metrics/tracing for decisions
-
{todo} Cover repo-mapping paths with tests and caching
-
{todo} Backpressure / retry strategy for downstream workflows
15. DataSource Integration
External data integration powers modern lending decisions through the DataSource framework and Feature Store. DataSources fetch raw data from external APIs, while the Feature Store transforms this data into usable features for business logic and decision workflows.
15.1. Architecture Overview
The TimveroOS data integration follows a three-layer architecture:
-
DataSource Layer - Fetches raw data from external APIs
-
DataSourceManager - Manages caching, loading modes, and data lifecycle
-
Feature Store - Transforms raw data into business features through configurable mappings
External API → DataSource → DataSourceManager → Feature Store → Business Logic
Key Principle: DataSources should never be used directly. Always access them through DataSourceManager or Feature Store.
15.2. DataSource Framework
Core Interfaces
public interface DataSource<E extends DataSourceSubject> {
Content getData(E subject) throws Exception;
Duration lifespan(); // How long data remains valid
class Content {
private final byte[] body;
private final String mime;
// getters...
}
}
public interface MappedDataSource<E, T> extends DataSource<E> {
Class<T> getType(); // Target parsing type
T parseRecord(Content data); // Parse raw data to typed object
}
DataSourceManager - The Proper Access Layer
Never use DataSources directly. Always use DataSourceManager which provides:
-
Intelligent caching - Avoids redundant API calls
-
Loading modes - Control when to fetch vs use cached data
-
Data lifecycle - Automatic expiration and invalidation
-
Error handling - Graceful degradation when data unavailable
public interface DataSourceManager {
<E extends DataSourceSubject> Optional<DataSourceRecord> getData(
E entity, String dataSourceName, LoadingMode mode) throws IOException;
enum LoadingMode {
READ, // Use cached data only
QUERY, // Use cache or fetch if missing
FORCE // Always fetch fresh data
}
}
15.3. Complete Implementation Example: GitHub DataSource
The GitHub DataSource demonstrates a production-ready implementation that fetches user data from the GitHub API for risk assessment purposes.
Service Declaration
@Service(GithubDataSource.DATASOURCE_NAME)
public class GithubDataSource implements MappedDataSource<GithubDataSourceSubject, GithubUser> {
public static final String DATASOURCE_NAME = "github";
private final RestTemplate restTemplate = new RestTemplate();
private final ObjectMapper objectMapper = new ObjectMapper();
private final String GITHUB_API_BASE_URL = "https://api.github.com";
Key Points:
-
@Servicewith name - Makes DataSource discoverable by the platform -
Constant naming - Consistent reference for DataSource identification
-
RestTemplate - Spring’s HTTP client for API calls
-
ObjectMapper - Jackson for JSON parsing with flexible configuration
HTTP Configuration
{
objectMapper.configure(FAIL_ON_UNKNOWN_PROPERTIES, false);
}
private HttpEntity<String> createHttpEntity() {
HttpHeaders headers = new HttpHeaders();
headers.set("Accept", "application/vnd.github.v3+json");
return new HttpEntity<>(headers);
}
Best Practices:
-
Flexible JSON parsing -
FAIL_ON_UNKNOWN_PROPERTIES = falsehandles API changes -
API versioning - Explicit version headers ensure consistent responses
-
Reusable headers - Centralized HTTP configuration
Data Retrieval Implementation
@Override
public Content getData(GithubDataSourceSubject subject) throws Exception {
try {
String url = GITHUB_API_BASE_URL + "/users/" + subject.getGithubUsername();
ResponseEntity<byte[]> response = restTemplate.exchange(
url,
HttpMethod.GET,
createHttpEntity(),
byte[].class
);
return new Content(response.getBody(), MediaType.APPLICATION_JSON_VALUE);
} catch (HttpClientErrorException.NotFound e) {
throw new DataUnavailableException("User not found: " + subject.getGithubUsername());
}
}
Implementation Details:
-
URL construction - Safe string concatenation with subject data
-
Byte array response - Preserves raw data for Content object
-
Exception mapping - HTTP 404 becomes
DataUnavailableException -
Media type preservation - Maintains content type for parsing
Type Information and Parsing
@Override
public Class<GithubUser> getType() {
return GithubUser.class;
}
@Override
public GithubUser parseRecord(Content data) throws IOException {
return objectMapper.readValue(data.getBody(), GithubUser.class);
}
Type Safety:
-
Generic type preservation -
getType()enables runtime type checking -
Automatic parsing - Platform can automatically parse Content to target type
-
Exception handling - Jackson exceptions bubble up as DataSource exceptions
15.4. Subject and Target Objects
Subject Interface
The subject defines what data to fetch from the external source:
public interface GithubDataSourceSubject {
String getGithubUsername();
}
Design Principles:
-
Interface-based - Allows multiple entities to implement the same subject
-
Minimal contract - Only required data for the external API call
-
Clear naming - Method names match the external API requirements
Target Data Class
The target object represents the parsed external data:
public class GithubUser {
private String login;
private String name;
@JsonProperty("followers")
private int followersCount;
@JsonProperty("following")
private int followingCount;
@JsonProperty("public_repos")
private int publicRepos;
@JsonProperty("avatar_url")
private String avatarUrl;
// Constructors, getters, and setters...
}
JSON Mapping:
-
@JsonProperty- Maps JSON field names to Java properties -
Selective fields - Only include relevant data for your application
-
Type safety - Strong typing for external API responses
15.5. Entity Integration Pattern
The power of DataSources comes from integrating them directly with your business entities:
Entity Implementation
@Entity
public class Participant extends AbstractAuditable<UUID>
implements GithubDataSourceSubject {
@Column
private String githubUsername;
@Override
public String getGithubUsername() {
return githubUsername;
}
// Other entity fields and methods...
}
Integration Benefits:
-
Direct entity support - No additional mapping layers needed
-
Type safety - Compile-time checking of subject contracts
-
Automatic discovery - Platform can find applicable DataSources
Proper Usage Through DataSourceManager
-
WRONG - Never use DataSources directly:
@Autowired
@Qualifier("github")
private MappedDataSource<GithubDataSourceSubject, GithubUser> githubDataSource;
// DON'T DO THIS - bypasses caching and lifecycle management
GithubUser data = githubDataSource.parseRecord(githubDataSource.getData(participant));
-
CORRECT - Use DataSourceManager:
@Autowired
private DataSourceManager dataSourceManager;
public void enrichParticipantData(Participant participant) {
try {
Optional<DataSourceRecord> record = dataSourceManager.getData(
participant, "github", LoadingMode.QUERY);
if (record.isPresent()) {
GithubUser githubData = (GithubUser) record.get().getData();
assessDeveloperRisk(participant, githubData);
}
} catch (IOException e) {
log.warn("GitHub data unavailable for participant: {}", participant.getId());
// Continue without GitHub data
}
}
Loading Mode Benefits:
-
READ- Fast, uses only cached data for performance-critical paths -
QUERY- Balanced, fetches if needed for standard workflows -
FORCE- Fresh data for critical decisions or data refresh workflows
15.6. Feature Store Integration
The Feature Store is the primary way to consume DataSource data in business logic. It transforms raw external data into structured features through configurable field mappings.
What are Features?
A feature is a data transformation that converts raw data from integrated sources into a format usable by workflow decision logic:
-
Direct value extractions - Credit scores from bureau data
-
Calculated values - Debt-to-income ratios
-
Derived indicators - Payment pattern analysis
-
Aggregated metrics - Total outstanding debt
Feature Store Benefits
-
Configurable transformations - Change feature extraction without code changes
-
Automatic caching - Features are computed once and stored
-
Version management - Track changes to feature definitions
-
Type safety - Features have defined data types
-
Error handling - Graceful handling of transformation failures
-
Audit trail - Complete history of feature values
-
Performance - Bulk feature extraction and caching
The Feature Store automatically uses DataSourceManager to fetch data with appropriate caching and lifecycle management, then applies configurable transformations to create business-ready features.
Note: Feature Store implementation and usage is covered in the Feature Store documentation. This chapter focuses on the underlying DataSource implementation that powers the Feature Store.
15.7. Advanced Patterns
Multiple DataSource Support
Entities can implement multiple subject interfaces for different data sources:
@Entity
public class Participant implements GithubDataSourceSubject, CreditBureauSubject {
@Override
public String getGithubUsername() {
return githubUsername;
}
@Override
public String getNationalId() {
return getClient().getIndividualInfo().getNationalId();
}
}
DataSource Lifespan Configuration
Configure how long data remains valid to balance freshness vs performance:
@Service("github")
public class GithubDataSource implements MappedDataSource<GithubDataSourceSubject, GithubUser> {
@Override
public Duration lifespan() {
return Duration.ofHours(24); // GitHub data valid for 24 hours
}
@Override
public Content getData(GithubDataSourceSubject subject) throws Exception {
// Implementation...
}
}
Error Recovery Strategies
Implement fallback mechanisms for critical data sources using DataSourceManager:
@Autowired
private DataSourceManager dataSourceManager;
public GithubUser getGithubDataWithFallback(Participant participant) {
try {
Optional<DataSourceRecord> record = dataSourceManager.getData(
participant, "github", LoadingMode.QUERY);
if (record.isPresent()) {
return (GithubUser) record.get().getData();
}
} catch (IOException e) {
log.warn("Primary GitHub data unavailable, trying fallback", e);
}
// Try with cached data only as fallback
try {
Optional<DataSourceRecord> cachedRecord = dataSourceManager.getData(
participant, "github", LoadingMode.READ);
if (cachedRecord.isPresent()) {
log.info("Using cached GitHub data for participant: {}", participant.getId());
return (GithubUser) cachedRecord.get().getData();
}
} catch (IOException e) {
log.warn("Cached GitHub data also unavailable", e);
}
return null; // No data available
}
15.8. Common Use Cases
Credit Bureau Integration
@Service("creditBureau")
public class CreditBureauDataSource
implements MappedDataSource<CreditBureauSubject, CreditReport> {
@Override
public Content getData(CreditBureauSubject subject) throws Exception {
// Call credit bureau API with SSN/National ID
// Handle authentication, rate limiting, etc.
}
}
15.9. Testing DataSources
Unit Testing
@Test
public void testGithubDataSource() throws Exception {
GithubDataSourceSubject subject = () -> "octocat";
Content content = githubDataSource.getData(subject);
GithubUser user = githubDataSource.parseRecord(content);
assertThat(user.getLogin()).isEqualTo("octocat");
assertThat(user.getPublicRepos()).isGreaterThan(0);
}
15.10. Best Practices
Architecture Patterns
-
Use Feature Store for business logic - Primary pattern for consuming external data
-
Use DataSourceManager for direct access - When you need raw data or custom processing
-
Never use DataSources directly - Always go through DataSourceManager or Feature Store
-
Choose appropriate loading modes - READ for performance, QUERY for balance, FORCE for freshness
-
Handle data unavailability gracefully - Continue workflow when external data is missing
-
Implement proper subject interfaces - Clear contracts for what data to fetch
-
Use typed target objects - Strong typing for external API responses
DataSource Implementation
-
Use meaningful service names -
@Service("github")not@Service("ds1") -
Handle errors gracefully - Always throw
DataUnavailableExceptionfor missing data -
Configure JSON parsing - Use
FAIL_ON_UNKNOWN_PROPERTIES = falsefor API resilience -
Set appropriate lifespans - Balance freshness vs API call costs
-
Version your APIs - Use explicit API version headers
-
Test thoroughly - Test both success and failure scenarios
-
Implement proper parsing - Handle all expected data formats and edge cases
DataSourceManager Usage
-
Use appropriate loading modes - Match mode to business requirements
-
Handle Optional results - Check if data is present before using
-
Catch IOException properly - Handle network and data access failures
-
Log data access patterns - Monitor usage for performance optimization
-
Implement fallback strategies - Use cached data when fresh data unavailable
Security and Performance
-
Log appropriately - Log errors but not sensitive data
-
Add retry logic - Handle temporary network failures
-
Set reasonable timeouts - Don’t block indefinitely
-
Monitor data freshness - Track when data was last updated
-
Use lifespans effectively - Avoid unnecessary API calls
Anti-Patterns
-
{bad} Don’t use DataSources directly - Bypasses caching and lifecycle management
-
{bad} Don’t bypass Feature Store for business logic - Use Feature Store instead of raw DataSource data
-
{bad} Don’t ignore Optional results - Always check if data is present
-
{bad} Don’t hardcode loading modes - Choose based on business requirements
-
{bad} Don’t expose sensitive data - Never log API keys or personal information
-
{bad} Don’t hardcode URLs - Use configuration properties for API endpoints
-
{bad} Don’t ignore exceptions - Handle
IOExceptionandDataUnavailableException
Production Checklist
-
{todo} DataSources use meaningful service names
-
{todo} All external calls have appropriate timeouts
-
{todo} Error handling covers
DataUnavailableExceptionandIOException -
{todo} Loading modes are chosen appropriately for each use case
-
{todo} Sensitive data is never logged
-
{todo} DataSource lifespans balance freshness vs cost
-
{todo} Subject interfaces are properly implemented
-
{todo} Target objects handle all expected data formats
16. Workflow Integration
Connect your entities to automated decision-making workflows for credit scoring, risk assessment, and approval processes.
16.1. What You’ll Learn
After reading this chapter, you’ll know how to:
-
Make any entity workflow-enabled with two simple interfaces
-
Automatically trigger workflows when business events occur
-
Handle workflow results and update entity status
-
Provide manual workflow controls for users
-
Test your workflow integration
16.2. The Big Picture
TimveroOS separates what decisions to make from how to make them:
-
Workflow Engine (admin configures): Decision logic, scoring rules, approval criteria
-
Your Code (this chapter): When to start workflows, how to handle results
Think of it like this: You tell the system "start credit check for this customer," the workflow engine figures out approve/decline/review, then you handle the result.
16.3. Step 1: Make Your Subject Workflow-Enabled
ProcessEntity represents the subject being evaluated - typically a person, property, or asset. Not the business process itself.
@Entity
public class Borrower extends AbstractAuditable<UUID>
implements ProcessEntity, HasPendingDecisions {
// Person/subject information
private String fullName;
private String socialSecurityNumber;
private String email;
private LocalDate dateOfBirth;
// Workflow integration - just add these two things:
@OneToOne(cascade = CascadeType.ALL)
private PendingDecisionHolder pendingDecisionHolder =
new PendingDecisionHolder("BORROWER");
@Override
public String getPrimaryId() {
return socialSecurityNumber; // Unique identifier for data sources
}
@Override
public PendingDecisionHolder getPendingDecisionHolder() {
return pendingDecisionHolder;
}
}
Common ProcessEntity types:
-
Person: Borrower, Guarantor, Co-signer (credit checks, income verification)
-
Property: House, Vehicle, Asset (appraisals, valuations)
-
Business: Company, Partnership (business credit, financial analysis)
What Each Interface Does
ProcessEntity: "This subject can be evaluated by workflows"
-
Requires
getPrimaryId()- how external data sources identify this subject (SSN, VIN, Tax ID, etc.)
HasPendingDecisions: "This subject can receive workflow evaluation results"
-
Stores decisions from workflows (approve/decline/manual review)
-
Tracks decision progress
16.4. Step 2: Automatically Start Workflows
Use EntityChecker to start workflows when business events happen. The pattern is to listen for business process changes and start subject evaluation:
@Component
public class BorrowerCreditCheckTrigger extends EntityChecker<LoanApplication, UUID> {
@Autowired
private DecisionProcessStarter workflowStarter;
@Override
protected void registerListeners(CheckerListenerRegistry<LoanApplication> registry) {
// Listen for application status changes
registry.entityChange().updated(LoanApplication_.STATUS);
}
@Override
protected boolean isAvailable(LoanApplication application) {
return application.getStatus() == ApplicationStatus.SUBMITTED
&& application.getBorrower() != null;
}
@Override
protected void perform(LoanApplication application) {
// Start credit check workflow for the BORROWER (the subject)
Borrower borrower = application.getBorrower();
workflowStarter.start(CREDIT_CHECK_WORKFLOW, borrower.getId());
}
}
Key pattern:
-
Listen for business process events (application submitted, document signed, etc.)
-
Check if subject evaluation should start (
isAvailable) -
Start workflow for the subject (borrower, property, etc.), not the process
Common Trigger Patterns
Status changes:
registry.entityChange().updated(MyEntity_.STATUS);
New records:
registry.entityChange().inserted();
Specific field updates:
registry.entityChange().updated(MyEntity_.CREDIT_SCORE);
Complex conditions:
registry.entityChange().updated(MyEntity_.STATUS)
.and(entity -> entity.getAmount().isGreaterThan(THRESHOLD));
Manual Workflow Calls
Sometimes you need to start workflows directly from your service code:
@Service
public class BorrowerEvaluationService {
@Autowired
private DecisionProcessStarter workflowStarter;
public void requestCreditCheck(UUID borrowerId) {
workflowStarter.start(CREDIT_CHECK_WORKFLOW, borrowerId);
}
public void requestIncomeVerification(UUID borrowerId) {
workflowStarter.start(INCOME_VERIFICATION_WORKFLOW, borrowerId);
}
public void evaluateAllBorrowersForApplication(UUID applicationId) {
LoanApplication app = applicationRepository.findById(applicationId);
for (Borrower borrower : app.getBorrowers()) {
workflowStarter.start(CREDIT_CHECK_WORKFLOW, borrower.getId());
}
}
}
Use this when:
-
User clicks "Run Credit Check" button for a specific borrower
-
External API triggers evaluation of a person/property
-
Scheduled job needs to re-evaluate subjects
-
You need precise control over which subjects to evaluate
16.5. Step 3: Handle Workflow Results
When workflows complete, they send results back to your subject. Handle them with EntityEventListener:
@Component
public class BorrowerEvaluationResultHandler implements EntityEventListener<FinishedScoringEvent<Borrower>> {
@Autowired
private BorrowerService borrowerService;
@Autowired
private LoanApplicationService applicationService;
@Override
@Transactional(propagation = Propagation.REQUIRES_NEW)
public void handle(FinishedScoringEvent<Borrower> event) {
UUID borrowerId = event.getEntityId();
Borrower borrower = borrowerService.findById(borrowerId);
List<PendingDecision> decisions = borrower.getPendingDecisions();
if (hasDeclinedDecisions(decisions)) {
borrowerService.markAsDeclined(borrowerId, getDeclineReason(decisions));
// Update related applications
applicationService.handleBorrowerDeclined(borrowerId);
} else if (hasPendingDecisions(decisions)) {
borrowerService.markForManualReview(borrowerId);
} else {
borrowerService.markAsApproved(borrowerId);
// Check if all borrowers are approved, then approve application
applicationService.checkApplicationReadiness(borrower.getApplicationId());
}
}
private boolean hasDeclinedDecisions(List<PendingDecision> decisions) {
return decisions.stream().anyMatch(d -> d.getStatus() == DecisionStatus.DECLINED);
}
private boolean hasPendingDecisions(List<PendingDecision> decisions) {
return decisions.stream().anyMatch(d -> d.getStatus() == DecisionStatus.PENDING);
}
}
What this does:
-
Listens for subject evaluation completion events
-
Checks all decisions from the workflow
-
Updates subject status based on results
-
Triggers business logic (e.g., check if application can proceed)
-
Uses new transaction to avoid conflicts with workflow engine
Service Layer Pattern
Keep your business logic clean with separate services for subjects and processes:
@Service
public class BorrowerService {
@Transactional
public void markAsApproved(UUID borrowerId) {
Borrower borrower = repository.findById(borrowerId);
borrower.setEvaluationStatus(EvaluationStatus.APPROVED);
borrower.setEvaluationDate(Instant.now());
// Clear any pending decisions, update credit score, etc.
}
@Transactional
public void markAsDeclined(UUID borrowerId, String reason) {
Borrower borrower = repository.findById(borrowerId);
borrower.setEvaluationStatus(EvaluationStatus.DECLINED);
borrower.setDeclineReason(reason);
// Log decline reason, update risk profile, etc.
}
@Transactional
public void markForManualReview(UUID borrowerId) {
Borrower borrower = repository.findById(borrowerId);
borrower.setEvaluationStatus(EvaluationStatus.MANUAL_REVIEW);
// Create review tasks, notify underwriters, etc.
}
}
@Service
public class LoanApplicationService {
@Transactional
public void checkApplicationReadiness(UUID applicationId) {
LoanApplication app = repository.findById(applicationId);
// Check if all borrowers are evaluated
boolean allBorrowersReady = app.getBorrowers().stream()
.allMatch(b -> b.getEvaluationStatus() != EvaluationStatus.PENDING);
if (allBorrowersReady) {
boolean anyDeclined = app.getBorrowers().stream()
.anyMatch(b -> b.getEvaluationStatus() == EvaluationStatus.DECLINED);
if (anyDeclined) {
app.setStatus(ApplicationStatus.DECLINED);
} else {
app.setStatus(ApplicationStatus.APPROVED);
}
}
}
}
16.6. Step 4: Add Manual Controls
Sometimes users need to manually control workflows. Add action controllers:
@Controller
@RequestMapping("/retry-credit-check")
public class RetryCreditCheckAction extends SimpleActionController<UUID, Borrower> {
@Autowired
private DecisionProcessStarter workflowStarter;
@Override
protected EntityAction<? super Borrower, Object> action() {
return when(borrower ->
borrower.getEvaluationStatus() == EvaluationStatus.FAILED
).then((borrower, form, user) -> {
workflowStarter.start(CREDIT_CHECK_WORKFLOW, borrower.getId());
borrower.setEvaluationStatus(EvaluationStatus.PENDING);
});
}
}
This creates a "Retry Credit Check" button that:
-
Only shows when borrower evaluation failed
-
Restarts the credit check workflow for that borrower
-
Updates borrower evaluation status
-
Refreshes the page
16.7. Step 5: Configuration
Define Your Workflow Types
Create DecisionProcessType constants for your workflows:
@Configuration
public class WorkflowConfiguration {
public static final DecisionProcessType<Borrower> CREDIT_CHECK_WORKFLOW =
new DecisionProcessType<>("CREDIT_CHECK_WORKFLOW", Borrower.class);
public static final DecisionProcessType<Borrower> INCOME_VERIFICATION_WORKFLOW =
new DecisionProcessType<>("INCOME_VERIFICATION_WORKFLOW", Borrower.class);
public static final DecisionProcessType<Borrower> FRAUD_CHECK_WORKFLOW =
new DecisionProcessType<>("FRAUD_CHECK_WORKFLOW", Borrower.class);
public static final DecisionProcessType<Property> PROPERTY_APPRAISAL_WORKFLOW =
new DecisionProcessType<>("PROPERTY_APPRAISAL_WORKFLOW", Property.class);
}
Each DecisionProcessType specifies:
-
Name: Identifier for the workflow process
-
Entity type: What type of entity this workflow processes
Use these constants everywhere instead of creating new instances.
Application Setup
The workflow engine runs on a separate port. Set this up in your main class:
public class MyLendingApplication {
public static void main(String[] args) {
SpringApplicationBuilder parent = new SpringApplicationBuilder(BaseConfiguration.class)
.web(WebApplicationType.NONE);
parent.run(args);
// Main application (port 8081)
parent.child(WebMvcConfig.class, MyConfiguration.class)
.properties("server.port=8081")
.run(args);
// Workflow engine (separate port)
parent.child(ExternalProcessWebMvcConfig.class)
.properties("spring.config.name=workflow")
.run(args);
}
}
Workflow Properties
Create src/main/resources/workflow.properties:
server.port=${process.engine.callbackPort}
server.servlet.context-path=/external-process
Application Properties
Add these essential workflow configuration properties to src/main/resources/application.properties:
# Workflow Callback Configuration
process.engine.callbackPort=8180
process.engine.callbackUrl=http://localhost:
process.engine.type=workflow
# Workflow Modeler UI
process.modeler.url=http://localhost:8280/workflow
# Workflow Engine URL for back-to-back calls
process.engine.url=http://localhost:8280/workflow
What these properties do:
-
process.engine.callbackPort: Port where your admin application runs (workflow engine calls back to this) -
process.engine.callbackUrl: Base URL for workflow engine callbacks to your application -
process.engine.type: Identifies this as a workflow-enabled application -
process.modeler.url: URL to the workflow designer/modeler interface -
process.engine.url: URL to the workflow execution engine
Important: The workflow engine runs separately from your application and needs these URLs to communicate back and forth.
16.8. Complete Example: Borrower Credit Check Workflow
Here’s everything working together for a borrower evaluation workflow:
1. The Subject (ProcessEntity)
@Entity
public class Borrower extends AbstractAuditable<UUID>
implements ProcessEntity, HasPendingDecisions {
private String fullName;
private String socialSecurityNumber;
private EvaluationStatus evaluationStatus = EvaluationStatus.PENDING;
@OneToOne(cascade = CascadeType.ALL)
private PendingDecisionHolder pendingDecisionHolder =
new PendingDecisionHolder("BORROWER");
@Override
public String getPrimaryId() { return socialSecurityNumber; }
@Override
public PendingDecisionHolder getPendingDecisionHolder() {
return pendingDecisionHolder;
}
// getters/setters...
}
2. Automatic Trigger (listens to business process)
@Component
public class BorrowerCreditCheckTrigger extends EntityChecker<LoanApplication, UUID> {
@Override
protected void registerListeners(CheckerListenerRegistry<LoanApplication> registry) {
registry.entityChange().updated(LoanApplication_.STATUS);
}
@Override
protected boolean isAvailable(LoanApplication app) {
return app.getStatus() == ApplicationStatus.SUBMITTED;
}
@Override
protected void perform(LoanApplication app) {
// Start workflow for each BORROWER (the subject)
for (Borrower borrower : app.getBorrowers()) {
workflowStarter.start(CREDIT_CHECK_WORKFLOW, borrower.getId());
}
app.setStatus(ApplicationStatus.UNDER_REVIEW);
}
}
3. Result Handler (handles subject evaluation results)
@Component
public class BorrowerEvaluationResultHandler implements EntityEventListener<FinishedScoringEvent<Borrower>> {
@Override
@Transactional(propagation = Propagation.REQUIRES_NEW)
public void handle(FinishedScoringEvent<Borrower> event) {
Borrower borrower = borrowerService.findById(event.getEntityId());
List<PendingDecision> decisions = borrower.getPendingDecisions();
if (hasDeclinedDecisions(decisions)) {
borrowerService.markAsDeclined(borrower.getId());
applicationService.handleBorrowerDeclined(borrower.getApplicationId());
} else if (hasPendingDecisions(decisions)) {
borrowerService.markForManualReview(borrower.getId());
} else {
borrowerService.markAsApproved(borrower.getId());
applicationService.checkApplicationReadiness(borrower.getApplicationId());
}
}
}
4. Manual Controls
@Controller
@RequestMapping("/retry-credit-check")
public class RetryCreditCheckAction extends SimpleActionController<UUID, Borrower> {
@Override
protected EntityAction<? super Borrower, Object> action() {
return when(borrower -> borrower.getEvaluationStatus() == EvaluationStatus.FAILED)
.then((borrower, form, user) -> {
workflowStarter.start(CREDIT_CHECK_WORKFLOW, borrower.getId());
borrower.setEvaluationStatus(EvaluationStatus.PENDING);
});
}
}
The Flow:
-
User submits application → Application status changes to SUBMITTED
-
EntityChecker detects change → Starts credit check workflow for each borrower
-
Workflow evaluates borrower → Sends result to your handler
-
Handler updates borrower status → Checks if application can proceed
-
If workflow fails → User can retry credit check for specific borrower
Key Pattern: Business process (application) triggers subject evaluation (borrower), results flow back to update both subject and process.
17. Product Web Components Configuration
Purpose: Automate the setup of web layer components (controllers, actions, tabs) for credit products and additives through declarative configuration.
The ProductWebConfigurationSupport class provides a declarative approach to registering all necessary Spring beans for credit product management, eliminating boilerplate code and ensuring consistent product UI implementation across your application.
17.1. Problem Statement
Building a complete product management UI typically requires:
-
Controllers for listing and viewing products
-
Actions for CRUD operations (create, edit, copy, activate, deactivate)
-
Additive actions for managing product variations
-
Tabs for product details and additives list
-
Proper generic type resolution for Spring dependency injection
17.2. What You’ll Build
To create a complete product management system, you need to implement:
Required Components (you implement these):
-
Product Entity: Your
CreditProductsubclass with business-specific fields -
Additive Entity: Your
CreditProductAdditivesubclass for product variations -
Form Classes: DTOs for product and additive creation/editing
-
Form Services: Services implementing
EntityFormServiceandCreditProductAdditiveFormService -
MapStruct Mappers: Bidirectional mapping between entities and forms
-
Thymeleaf Templates: UI templates for forms, lists, and tabs
-
Filter Class:
BaseCreditProductFiltersubclass for list filtering -
Configuration Class: Single class extending
ProductWebConfigurationSupport
Auto-Generated Components (framework provides these):
-
Controller:
CreditProductControllerfor list and details pages -
12+ Action Beans: CRUD operations (create, edit, copy, activate, deactivate, additive management)
-
Tab Components: Product details and additives tabs
-
Type-Safe Wiring: Proper generic type resolution for dependency injection
17.3. Core Concepts
ProductWebConfigurationSupport
Abstract base class that implements BeanDefinitionRegistryPostProcessor to programmatically register Spring beans for credit product web layer.
Key Features:
-
Automatic Bean Registration: Creates all necessary controllers, actions, and tabs
-
Generic Type Resolution: Ensures proper dependency injection with generic parameters
-
Template Customization: Override methods to specify custom Thymeleaf template paths
-
Extensible Design: Add custom components by overriding
initDefinitions()
Product Lifecycle States
Credit products follow a three-state lifecycle:
DRAFT:
-
Product is being configured and not yet ready for use
-
Full edit access - all product fields and properties can be modified
-
Additives can be created, edited, activated, and deactivated
-
Not available for offer generation
-
Can be activated when at least one active additive exists
-
Can be copied to create a new DRAFT product
ACTIVE:
-
Product is live and available for offer generation
-
Product configuration is locked - no editing of existing fields
-
Additives cannot be created, edited, or modified
-
Can be deactivated (moved to INACTIVE state)
-
Can be copied to create a new DRAFT product with all active additives
INACTIVE:
-
Product has been deactivated (soft delete)
-
Not available for offer generation
-
Product remains in database for historical reference
-
Cannot be edited or reactivated through standard actions
-
Can be copied to create a new DRAFT product
Standard Components Registered
When you extend ProductWebConfigurationSupport, the following beans are automatically registered:
| Component Type | Bean Class | Purpose |
|---|---|---|
Controller |
|
List view and product details display |
Product Actions |
|
Create new credit product |
|
Edit existing product |
|
|
Copy product with all additives |
|
|
Deactivate product (soft delete) |
|
|
Activate/enable product |
|
|
Add new additive to product |
|
Additive Actions |
|
Edit product additive |
|
Activate additive |
|
|
Deactivate additive |
|
Tabs |
|
Product details tab view |
|
Product additives tab view |
17.4. Complete Implementation Checklist
Follow these steps to implement all required components for your product management system.
✓ Step 1: Product Entity
Create your CreditProduct subclass with business-specific fields
@Entity
public class ExampleCreditProduct extends CreditProduct {
private CurrencyUnit currency;
private BigDecimal minAmount;
private BigDecimal maxAmount;
private Integer minTerm;
// ... other fields (maxTerm, lateFeeRate, etc.)
}
| See complete example: ExampleCreditProduct.java |
✓ Step 2: Additive Entity
Create your CreditProductAdditive subclass for product variations
@Entity
public class ExampleCreditProductAdditive extends CreditProductAdditive {
private String name;
private BigDecimal interestRate;
private BigDecimal minAmount;
// ... other fields (maxAmount, minTerm, maxTerm, etc.)
@Override
public ExampleProductOffer createOffer() {
return new ExampleProductOffer();
}
}
| See complete example: ExampleCreditProductAdditive.java |
✓ Step 3: Product Form Class
Create form DTO for product creation/editing
public class CreditProductForm extends BaseCreditProductForm {
private CurrencyUnit currency;
private BigDecimal minAmount;
private BigDecimal maxAmount;
// ... validation annotations and other fields
}
| See complete example: CreditProductForm.java |
✓ Step 4: Additive Form Class
Create form DTO for additive creation/editing
public class ExampleCreditProductAdditiveForm {
private String name;
private BigDecimal interestRate;
private BigDecimal minAmount;
// ... validation annotations and other fields
}
| See complete example: ExampleCreditProductAdditiveForm.java |
✓ Step 5: Filter Class
Create filter class for product list filtering
public class ExampleCreditProductFilter extends BaseCreditProductFilter {
}
Default Filters (from BaseCreditProductFilter ):
-
state: Filter products by state (
CreditProductState[]) - supports multiple states (IN restriction)
You can extend with custom filters using annotations:
-
@Field/@Fields- specify which entity fields to search -
@Restriction- define filter type (EQ, IN, LIKE, GT, LT, etc.)
| See complete example: ExampleCreditProductFilter.java |
✓ Step 6: MapStruct Mappers
Create MapStruct mappers for entity ↔ form conversion
Product Mapper:
@Mapper(unmappedTargetPolicy = ReportingPolicy.IGNORE)
public interface ExampleCreditProductFormMapper
extends EntityToFormMapper<ExampleCreditProduct, CreditProductForm> {
@Override
void toEntity(CreditProductForm form, @MappingTarget ExampleCreditProduct entity);
@Override
@InheritInverseConfiguration(name = "toEntity")
CreditProductForm toForm(ExampleCreditProduct entity);
@Override
@InheritConfiguration
ExampleCreditProduct createEntity(CreditProductForm form);
}
Additive Mapper:
@Mapper(uses = ReferenceMapper.class)
public interface ExampleCreditProductAdditiveFormMapper
extends EntityToFormMapper<ExampleCreditProductAdditive, ExampleCreditProductAdditiveForm> {
@Override
void toEntity(ExampleCreditProductAdditiveForm form,
@MappingTarget ExampleCreditProductAdditive entity);
@Override
@InheritInverseConfiguration(name = "toEntity")
ExampleCreditProductAdditiveForm toForm(ExampleCreditProductAdditive entity);
// Optional: custom mapping logic
@AfterMapping
default void afterToForm(@MappingTarget ExampleCreditProductAdditiveForm form,
ExampleCreditProductAdditive additive) {
form.setAdditiveId(additive.getId());
}
}
| See complete examples: ExampleCreditProductFormMapper.java, ExampleCreditProductAdditiveFormMapper.java |
✓ Step 7: Product Form Service
Implement form service for product operations
@Service
public class ExampleCreditProductFormService
extends EntityFormService<ExampleCreditProduct, CreditProductForm, UUID> {
@Autowired
private DocumentTemplateFormService documentTemplateService;
@Override
protected void assembleEditModel(ExampleCreditProduct entity,
CreditProductForm form,
Map<String, Object> model) {
// Add data needed for form rendering (dropdowns, selects, etc.)
model.put("productEngines", SimpleScheduledEngine.NAME);
model.put("offerEngineTypes", new ExecutionResultType[]{ExampleDataProcessor.TYPE});
model.put("creditTypes", CreditType.values(CreditType.class));
model.put("templates", documentTemplateService.getTemplatesMap(
ApplicationContractDocumentCategory.TYPE));
}
}
Key Method: assembleEditModel()
This method populates the model with data needed for form rendering:
-
Dropdown Options: Available engines, credit types, templates
-
Reference Data: Any lookup data that forms need (enums, lists, maps)
-
UI Data: Additional information for conditional rendering
| See complete example: ExampleCreditProductFormService.java |
✓ Step 8: Additive Form Service
Implement form service for additive operations
@Service
public class ExampleCreditProductAdditiveFormServiceImpl
extends CreditProductAdditiveFormService<
ExampleCreditProductAdditive,
ExampleCreditProductAdditiveForm> {
private final OfferEngineDescriptorRepository offerEngineDescriptorRepository;
@Override
public Collection<?> findAdditiveModelsByProduct(CreditProduct product) {
return product.getAdditives();
}
@Override
protected void assembleEditModel(@Nullable ExampleCreditProductAdditive entity,
ExampleCreditProductAdditiveForm form,
Map<String, Object> model) {
// Add data for additive form rendering
model.put("procuringTypes", ProcuringType.values());
}
@Override
public void assembleProductData(Model model, CreditProduct product) {
// Prepare data for additives tab
ExampleCreditProduct exampleProduct = (ExampleCreditProduct) product;
Set<ExecutionResultType> requiredEngineTypes = exampleProduct.getOfferEngineTypes();
model.addAttribute("productId", product.getId());
model.addAttribute("offerEngineTypes", requiredEngineTypes);
model.addAttribute("offerEngineDescriptors",
Lazy.of(() -> getEngineDescriptorMap(requiredEngineTypes)));
}
}
Key Methods:
-
assembleEditModel(): Populates model with data for additive edit form (procuring types dropdown)
-
assembleProductData(): Provides product-specific data for additives tab (engines, descriptors)
-
findAdditiveModelsByProduct(): Retrieves all additives for a product
| See complete example: ExampleCreditProductAdditiveFormServiceImpl.java |
✓ Step 9: Configuration Class
Create configuration class - this triggers automatic bean registration
@Configuration
@ConditionalOnWebApplication
public class ProductWebConfiguration extends ProductWebConfigurationSupport {
public ProductWebConfiguration() {
super(ExampleCreditProduct.class,
CreditProductForm.class,
CreditProductFilter.class,
ExampleCreditProductAdditive.class,
ExampleCreditProductAdditiveForm.class);
}
}
| See complete example: ProductWebConfiguration.java |
✓ Step 10: Thymeleaf Templates
Create UI templates for forms, lists, and tabs
You need to create templates at these paths (or override the paths in your configuration):
| Template Path | Purpose | Example |
|---|---|---|
|
Product list page with filters and actions |
|
|
Product create/edit form |
|
|
Product copy form |
Standard implementation provided |
|
Product details tab |
|
|
Product additives tab |
|
|
Add/edit additive form |
Template Structure Example:
src/main/resources/templates/
product/
list.html
edit.html
copy.html # Standard implementation provided
tab/
details.html
additives.html
action/
add-product-additive.html
The copy.html template has a standard implementation provided by the framework. You don’t need to create your own unless you require custom copy functionality beyond the default behavior.
|
| Review the example templates to understand the standard structure, form bindings, and integration with the framework’s UI components. |
✓ Step 11: Product State Label (Optional)
Create a label component to display product states in the UI
To show product states as visual labels in list views and detail pages, create a state label class:
@Component
@Order(1000)
public class CreditProductStateLabel extends EntityStatusLabel<ExampleCreditProduct> {
public CreditProductStateLabel() {
super(ExampleCreditProduct::getState);
}
@Override
public boolean isEntityMarked(ExampleCreditProduct entity) {
return entity.getState() != null;
}
@Override
public String getGroup() {
return "state";
}
}
Key Points:
-
Extends
EntityStatusLabelto automatically render state as a label -
Define translations in your message files:
productStateLabel.DRAFT,productStateLabel.ACTIVE,productStateLabel.INACTIVE
| See complete example: CreditProductStateLabel.java |
17.5. Template Customization
ProductWebConfigurationSupport provides default template paths, but you can customize them by overriding template path methods:
Available Template Path Methods
| Method | Default Path | Purpose |
|---|---|---|
|
|
Product list view |
|
|
Product details tab |
|
|
Product edit form |
|
|
Product creation form |
|
|
Product copy form |
|
|
Add additive form |
|
|
Edit additive form |
|
|
Additives tab view |
Customizing Templates
Override template path methods to use custom layouts:
@Configuration
@ConditionalOnWebApplication
public class CustomProductWebConfiguration extends ProductWebConfigurationSupport {
public CustomProductWebConfiguration() {
super(MyProduct.class, MyProductForm.class, MyProductFilter.class,
MyProductAdditive.class, MyAdditiveForm.class);
}
@Override
protected String getListTemplatePath() {
return "/custom/product/list-view";
}
@Override
protected String getEditTemplatePath() {
return "/custom/product/edit-form";
}
@Override
protected String getAdditivesTabTemplatePath() {
return "/custom/product/additives-panel";
}
// Override other template paths as needed
}
17.6. Standard Component Details
Product Actions
EditProductAction
Purpose: Edit existing product
Availability: Only for products in DRAFT state
Once a product is activated, it cannot be edited through this action.
CopyProductAction
Purpose: Duplicate product with all active additives
Availability: Always available for any product
Behavior:
* Copies product properties (code and title can be modified during copy)
* Duplicates only active additives
* New product starts in DRAFT state
EnableCreditProductAction
Purpose: Activate product for use in offer generation
Availability: Product must meet both conditions:
* State is DRAFT
* Product has at least one additive
Behavior: Changes product state from DRAFT to ACTIVE, making it available for offer generation.
Additive Actions
EditAdditiveAction
Purpose: Edit product additive
Availability: Only when product is in DRAFT state
Additives can only be edited while their parent product is in draft mode.
17.7. Next Steps
-
[offer-engine]: Use products to generate personalized offers
-
[credit-management]: Convert offers to active credits
-
[form-classes]: Advanced form handling and validation
-
[entity-checkers]: Automate product lifecycle events
18. Offer Engine & Credit Products
Purpose: Generate personalized loan offers based on credit products, participant data, and business rules.
The Offer Engine transforms static credit product templates into personalized offers by evaluating participant risk data, workflow results, and business rules through configurable scripts.
18.1. Process Flow
The offer generation follows this sequence:
| Product Template | Risk Assessment | Personalized Offer | Select Condition | Active Credit |
|---|---|---|---|---|
• Amount ranges • Interest rates • Terms • Payment engine |
• Credit score • Income data • Workflow results • External data |
• Adjusted amounts • Final rates • Secured options |
• Participant chooses offer • Payment schedule • Signs contract |
• Credit entity • Active servicing |
18.2. What You’ll Build
-
Credit Products: Loan product templates with terms and parameters
-
Product Additives: Specific configurations within products (rates, conditions)
-
Offer Generation: Automated personalized offer creation
-
Secured Offers: Collateral-based offer variants
-
Integration: Workflow-driven offer generation
18.3. Core Concepts
Credit Products
Credit products define the basic parameters for loan types - amounts, terms, currencies, and fees.
@Entity
public class ExampleCreditProduct extends CreditProduct {
private CurrencyUnit currency; // USD, EUR, etc.
private BigDecimal minAmount; // $1,000
private BigDecimal maxAmount; // $50,000
private Integer minTerm; // 6 months
private Integer maxTerm; // 60 months
private BigDecimal lateFeeRate; // 5% late fee
private String engineName; // "SimpleScheduledEngine"
// Constructor and getters/setters omitted for brevity
}
Key Features:
-
Amount Ranges: Min/max loan amounts with currency
-
Term Ranges: Min/max loan terms in months
-
Fee Structure: Late fee rates and other charges
-
Engine Integration: Links to offer generation engines
Payment Calculation Engines
Credit products integrate with payment calculation engines to generate detailed payment schedules. The engineName field references a Spring service that implements the ScheduledEngine interface to calculate how loan amounts are split into principal and interest payments over time.
// Example: SimpleScheduledEngine for standard annuity payments
@Service("SimpleScheduledEngine")
public class SimpleScheduledEngine implements ScheduledEngine<ExampleCreditCondition> {
@Override
public List<PaymentSegment> payments(ExampleCreditCondition condition,
LocalDate interestStart, LocalDate paymentStart,
MonetaryAmount principal, MonetaryAmount interest) {
// Calculate payment schedule based on loan terms
// Each PaymentSegment contains payment date, principal/interest breakdown,
// and remaining balances after payment
return payments;
}
}
Engine Purpose:
-
Payment Breakdown: Calculates how much of each payment goes to principal vs interest
-
Schedule Generation: Creates complete payment timeline with remaining balances
-
Algorithm Flexibility: Different engines can implement various payment calculation methods (annuity, differentiated, interest-only, etc.)
Integration:
-
Product Configuration: Credit product’s
engineNamefield matches Spring service name -
Automatic Discovery: Platform automatically finds and manages all ScheduledEngine beans
-
Validation: System ensures payment calculations are mathematically correct (remaining debt = 0)
Product Additives
Additives are specific configurations within products that define interest rates, procuring types, and offer generation rules.
@Entity
public class ExampleCreditProductAdditive extends CreditProductAdditive {
private String name; // "Prime Rate", "Standard Rate"
private BigDecimal interestRate; // 8.5%, 12.9%
private BigDecimal minAmount; // Refined amount ranges
private BigDecimal maxAmount; // within product limits
private Integer minTerm; // Refined term ranges
private Integer maxTerm; // within product limits
@Override
public ExampleProductOffer createOffer() {
return new ExampleProductOffer();
}
// Standard getters/setters omitted for brevity
}
Additive Purpose:
-
Interest Rates: Specific rates for different risk segments
-
Procuring Types: Collateral or guarantee requirements
-
Offer Engines: Scripts that generate personalized offers
-
Term Variations: Different terms within product ranges
Product Offers
Generated offers are personalized versions of product additives tailored to specific participants.
@Entity
public class ExampleProductOffer extends ProductOffer {
private UUID uuid; // Unique offer identifier
private Participant participant; // Who gets this offer
private BigDecimal minAmount; // Personalized min amount
private BigDecimal maxAmount; // Personalized max amount
private Integer minTerm; // Personalized min term
private Integer maxTerm; // Personalized max term
// Links back to product and participant's application
public Application getApplication() {
return participant.getApplication();
}
// Standard getters/setters omitted for brevity
}
Offer Characteristics:
-
Personalized Terms: Adjusted amounts and terms based on risk assessment
-
Participant Link: Connected to specific application participant
-
Product Reference: Links back to originating additive and product
-
UUID Tracking: Unique identifier for offer selection and tracking
18.4. Offer Generation Process
Core Engine
The OfferEngine orchestrates the generation process by combining product configurations with participant data.
Generation Flow:
-
Product Selection: Find products matching participant’s execution result type
-
Data Processing: Extract participant data using configured processor
-
Script Execution: Run offer generation scripts for each additive
-
Offer Creation: Generate personalized offers based on script results
-
Persistence: Save offers and link to participant
Data Processing
Implement OfferEngineDataProcessor to map participant data for offer generation:
Unresolved directive in _offer-engine.adoc - include::/opt/teamcity-agent3/work/8187ccecf864e5bd/src/main/java/com/timvero/example/admin/com/timvero/example/admin/offer/ExampleDataProcessor.java[tags=processor]
Data Sources:
-
Workflow Results: Decision process outcomes and scoring
-
Pending Decisions: Incomplete workflow data
-
Participant Profile: Personal and financial information
-
Risk Assessment: External data source results
Offer Service
The service layer coordinates offer generation with error handling:
Unresolved directive in _offer-engine.adoc - include::/opt/teamcity-agent3/work/8187ccecf864e5bd/src/main/java/com/timvero/example/admin/com/timvero/example/admin/offer/ProductOfferService.java[tags=service]
Service Responsibilities:
-
Transaction Management: Ensures atomic offer generation
-
Error Handling: Captures and stores generation exceptions
-
Product Loading: Retrieves active products for generation
-
Offer Persistence: Saves generated offers to database
18.5. Secured Offers & Procuring Engines
Secured offers extend basic offers with collateral or guarantee requirements through the ProcuringEngine pattern.
Procuring Engine Pattern
The ProcuringEngine transforms basic product offers into secured variants with specific collateral requirements:
@Component
public class PenaltyProcuringEngine implements ProcuringEngine {
@Override
public ProcuringType procuringType() {
return PENALTY;
}
@Override
public Collection<? extends SecuredOffer> generateSecuredOffers(ProductOffer productOffer) {
return List.of(new PenaltySecuredOffer((ExampleProductOffer) productOffer));
}
}
Engine Responsibilities:
-
Procuring Type: Defines what type of collateral/guarantee required
-
Offer Generation: Creates secured variants from basic offers
-
Business Logic: Implements specific procuring strategies
-
Flexibility: Multiple engines can handle different collateral types
Procuring Types
Define the types of collateral or guarantee requirements:
@Configuration
public class ExampleProcuringType {
public static final String CODE_PENALTY = "PENALTY";
public static final ProcuringType PENALTY = new ProcuringType(CODE_PENALTY);
// Other procuring types:
// VEHICLE_COLLATERAL, PROPERTY_COLLATERAL, COSIGNER, etc.
}
Common Procuring Types:
-
NO_PROCURING: Unsecured loans based on creditworthiness
-
PENALTY: Higher rates with penalty clauses for default
-
VEHICLE_COLLATERAL: Car loans with vehicle as collateral
-
PROPERTY_COLLATERAL: Mortgages with property as collateral
-
COSIGNER: Loans requiring guarantor/co-signer
Secured Offer Structure
Secured offers link to original offers while adding procuring-specific terms:
@Entity
@Table(name = "penalty_secured_offer")
@DiscriminatorValue(ExampleProcuringType.CODE_PENALTY)
public class PenaltySecuredOffer extends ExampleSecuredOffer {
protected PenaltySecuredOffer() {
}
public PenaltySecuredOffer(ExampleProductOffer originalOffer) {
super(originalOffer, ExampleProcuringType.PENALTY);
}
@Override
public String getOfferKey() {
return getOriginalOffer().getUuid() + ":PENALTY";
}
}
Base Secured Offer:
@Entity
@Table(name = "secured_offer")
@Inheritance(strategy = InheritanceType.JOINED)
@DiscriminatorColumn(name = "procuring_type", discriminatorType = DiscriminatorType.STRING)
public abstract class ExampleSecuredOffer extends SecuredOffer {
@ManyToOne(fetch = FetchType.EAGER, optional = false)
private ExampleProductOffer originalOffer;
@Column(name = "procuring_type", insertable = false, updatable = false)
private ProcuringType procuringType;
public ExampleSecuredOffer(ExampleProductOffer originalOffer, ProcuringType procuringType) {
this.originalOffer = originalOffer;
this.procuringType = procuringType;
}
// Abstract method for unique offer identification
public abstract String getOfferKey();
}
Secured Offer Features:
-
Original Offer Reference: Links to base product offer terms
-
Procuring Type: Specifies collateral/guarantee requirements
-
Inheritance Strategy: Supports multiple secured offer types
-
Offer Key: Unique identifier combining offer UUID and procuring type
-
Flexible Structure: Each procuring type can have custom fields and logic
Procuring Engine Integration
Procuring engines integrate with the main offer generation flow:
Generation Process:
-
Basic Offers: OfferEngine generates standard ProductOffers
-
Procuring Analysis: System identifies applicable procuring types
-
Secured Generation: Each ProcuringEngine creates secured variants
-
Offer Portfolio: Participant receives both basic and secured options
-
Selection: Participant chooses preferred offer type
Business Benefits:
-
Risk Mitigation: Collateral reduces lender risk
-
Rate Optimization: Secured offers can have lower interest rates
-
Market Expansion: Serve customers who need collateral-based options
-
Regulatory Compliance: Meet requirements for different loan types
Custom Procuring Engines
1. Define Procuring Type
@Configuration
public class MyProcuringTypes {
public static final String CODE_VEHICLE = "VEHICLE_COLLATERAL";
public static final ProcuringType VEHICLE = new ProcuringType(CODE_VEHICLE);
}
2. Implement Procuring Engine
@Component
public class VehicleProcuringEngine implements ProcuringEngine {
@Override
public ProcuringType procuringType() {
return MyProcuringTypes.VEHICLE;
}
@Override
public Collection<? extends SecuredOffer> generateSecuredOffers(ProductOffer productOffer) {
// Business logic for vehicle-secured offers
VehicleSecuredOffer securedOffer = new VehicleSecuredOffer((MyProductOffer) productOffer);
// Adjust terms based on vehicle value, age, etc.
return List.of(securedOffer);
}
}
3. Create Secured Offer Entity
@Entity
@Table(name = "vehicle_secured_offer")
@DiscriminatorValue(MyProcuringTypes.CODE_VEHICLE)
public class VehicleSecuredOffer extends MySecuredOffer {
@Column(name = "vehicle_value")
private BigDecimal vehicleValue;
@Column(name = "vehicle_year")
private Integer vehicleYear;
// Vehicle-specific offer logic
}
18.6. Integration Patterns
Offer to Credit Conversion
The conversion from offer to active credit happens in two phases: condition selection and contract signature.
Phase 1: Offer Selection & Condition Creation
When a participant selects an offer, the system creates a credit condition and prepares for contract signature:
@Controller
@RequestMapping("/submit-regular")
public class SelectRegularConditionAction extends SelectConditionAction<ExampleProductOffer, ConditionForm> {
@Override
protected EntityAction<? super ExampleProductOffer, ConditionForm> action() {
return when(o -> o.getApplication().getCondition() == null
&& o.getApplication().getStatus().equals(ApplicationStatus.CONDITION_CHOOSING)
&& o.getParticipant().getStatus() == ParticipantStatus.APPROVED).then((offer, form, user) -> {
Application application = offer.getApplication();
ExampleSecuredOffer securedOffer = findSecuredOffer(offer, form.getSecuredOfferKey());
// Calculate payment terms
MonetaryAmount principal = form.getPrincipal();
BigDecimal interestRate = offer.getProductAdditive().getInterestRate();
Integer term = form.getTerm();
MonetaryAmount regularPayment = PaymentCalculator.calcAnnuityPayment(
principal, MonetaryUtil.zero(principal.getCurrency()),
periodicInterest(Period.ofMonths(1), interestRate), term, 0);
// Create credit condition
ExampleCreditCondition condition = new ExampleCreditCondition(
principal, offer.getCreditProduct().getEngineName(),
interestRate, offer.getCreditProduct().getLateFeeRate(),
Method_30_360_BB.NAME, Period.ofMonths(1), term,
regularPayment, securedOffer);
application.setCondition(condition);
// Generate payment schedule
PaymentSchedule paymentSchedule = scheduledService.getPaymentSchedule(
condition, form.getPrincipal(), form.getStart());
application.setPaymentSchedule(paymentSchedule);
// Move to contract signature phase
application.setStatus(ApplicationStatus.PENDING_CONTRACT_SIGNATURE);
// Generate contract document
documentService.generate(application.getBorrowerParticipant(),
ParticipantDocumentTypesConfiguration.APPLICATION_CONTRACT,
offer.getCreditProduct().getUuidContractTemplate());
});
}
}
Phase 2: Contract Signature & Credit Creation
When the contract is signed, an EntityChecker automatically creates the active credit:
@Component
public class ContractSignChecker extends EntityChecker<Application, UUID> {
@Override
protected void registerListeners(CheckerListenerRegistry<Application> registry) {
registry.entityChange(SignableDocument.class,
d -> participantRepository.getReferenceById(d.getOwnerId()).getApplication())
.updated(SignableDocument_.STATUS)
.and(d -> d.getStatus() == SignatureStatus.SIGNED
&& d.getDocumentType() == ParticipantDocumentTypesConfiguration.APPLICATION_CONTRACT);
}
@Override
protected boolean isAvailable(Application application) {
return application.getStatus().equals(ApplicationStatus.PENDING_CONTRACT_SIGNATURE);
}
@Override
protected void perform(Application application) {
// Update application status
application.setStatus(ApplicationStatus.SERVICING);
// Get contract signature date
LocalDate signDate = documentFinder
.latest(application.getBorrowerParticipant(),
ParticipantDocumentTypesConfiguration.APPLICATION_CONTRACT)
.get().getDecisionMadeAt().atZone(ZoneId.systemDefault()).toLocalDate();
// Create active credit
ExampleCredit credit = new ExampleCredit();
credit.setApplication(application);
credit.setCondition(application.getCondition());
credit.setStartDate(signDate);
entityManager.persist(credit);
// Initialize credit calculations
calculationService.calculate(credit.getId(), signDate, signDate);
}
}
Complete Conversion Flow:
-
Offer Generation: Participant gets approved → offers generated via
GenerateOffersParticipantAction -
Offer Selection: Participant selects offer →
SelectRegularConditionActioncreatesCreditCondition -
Contract Generation: System generates contract document for signature
-
Contract Signature: Participant signs contract → triggers
ContractSignChecker -
Credit Creation: Checker creates
ExampleCreditand initializes calculations -
Servicing: Credit becomes active and ready for operations
Key Components:
-
ConditionForm: Captures participant’s chosen amount, term, and start date
-
CreditCondition: Immutable terms including payment calculation and secured offer reference
-
PaymentSchedule: Pre-calculated payment schedule based on selected terms
-
EntityChecker: Automated credit creation triggered by contract signature
-
CreditCalculationService: Initializes credit balances and schedules
Workflow Integration
Offers are typically generated after participant approval:
Unresolved directive in _offer-engine.adoc - include::/opt/teamcity-agent3/work/8187ccecf864e5bd/src/main/java/com/timvero/example/admin/com/timvero/example/admin/participant/action/GenerateOffersParticipantAction.java[tags=action]
Integration Points:
-
Status Checking: Only approved participants get offers
-
Error Recovery: Regenerate offers if previous attempt failed
-
Timing Control: Manual or automated generation triggers
-
UI Integration: Action buttons and status indicators
Display Integration
Format offers for user interface display:
Unresolved directive in _offer-engine.adoc - include::/opt/teamcity-agent3/work/8187ccecf864e5bd/src/main/java/com/timvero/example/admin/com/timvero/example/admin/application/ExampleProductOfferViewService.java[tags=view]
Display Features:
-
Localization: Multi-language offer descriptions
-
Formatting: Monetary amounts and interest rates
-
Details: Product names and procuring type descriptions
-
Comparison: Consistent format for offer comparison
18.7. Implementation Guide
Creating Credit Products
1. Define Product Entity
@Entity
public class MyLoanProduct extends CreditProduct {
// Custom fields for your loan type
private BigDecimal originationFee;
private Integer gracePeriodDays;
private String collateralRequirement;
// Constructor and getters/setters omitted for brevity
}
2. Configure Product Additives
@Entity
public class MyLoanAdditive extends CreditProductAdditive {
private String riskSegment;
private BigDecimal baseRate;
private Boolean allowsSecuredOffers;
@Override
public ProductOffer createOffer() {
return new MyLoanOffer();
}
// Getters/setters omitted for brevity
}
3. Implement Custom Offers
@Entity
public class MyLoanOffer extends ProductOffer {
private BigDecimal finalRate;
private String approvalConditions;
private BigDecimal loanToValueRatio;
// Getters/setters omitted for brevity
}
Implementing Secured Offers
1. Define Procuring Types
@Configuration
public class MyProcuringTypes {
public static final String CODE_VEHICLE = "VEHICLE_COLLATERAL";
public static final String CODE_PROPERTY = "PROPERTY_COLLATERAL";
public static final String CODE_COSIGNER = "COSIGNER";
public static final ProcuringType VEHICLE = new ProcuringType(CODE_VEHICLE);
public static final ProcuringType PROPERTY = new ProcuringType(CODE_PROPERTY);
public static final ProcuringType COSIGNER = new ProcuringType(CODE_COSIGNER);
}
2. Create Base Secured Offer
@Entity
@Table(name = "my_secured_offer")
@Inheritance(strategy = InheritanceType.JOINED)
@DiscriminatorColumn(name = "procuring_type", discriminatorType = DiscriminatorType.STRING)
public abstract class MySecuredOffer extends SecuredOffer {
@ManyToOne(fetch = FetchType.EAGER, optional = false)
private MyLoanOffer originalOffer;
@Column(name = "procuring_type", insertable = false, updatable = false)
private ProcuringType procuringType;
@Column(name = "adjusted_interest_rate")
private BigDecimal adjustedInterestRate;
protected MySecuredOffer() {}
public MySecuredOffer(MyLoanOffer originalOffer, ProcuringType procuringType) {
this.originalOffer = originalOffer;
this.procuringType = procuringType;
}
public abstract String getOfferKey();
// Getters/setters omitted for brevity
}
3. Implement Specific Secured Offers
@Entity
@Table(name = "vehicle_secured_offer")
@DiscriminatorValue(MyProcuringTypes.CODE_VEHICLE)
public class VehicleSecuredOffer extends MySecuredOffer {
@Column(name = "vehicle_value")
private BigDecimal vehicleValue;
@Column(name = "vehicle_year")
private Integer vehicleYear;
@Column(name = "vehicle_make")
private String vehicleMake;
@Column(name = "loan_to_value_ratio")
private BigDecimal loanToValueRatio;
protected VehicleSecuredOffer() {}
public VehicleSecuredOffer(MyLoanOffer originalOffer) {
super(originalOffer, MyProcuringTypes.VEHICLE);
}
@Override
public String getOfferKey() {
return getOriginalOffer().getUuid() + ":VEHICLE";
}
// Business logic methods
public boolean isEligibleVehicle() {
return vehicleYear >= 2015 && vehicleValue.compareTo(BigDecimal.valueOf(5000)) >= 0;
}
// Getters/setters omitted for brevity
}
@Entity
@Table(name = "cosigner_secured_offer")
@DiscriminatorValue(MyProcuringTypes.CODE_COSIGNER)
public class CosignerSecuredOffer extends MySecuredOffer {
@Column(name = "cosigner_credit_score")
private Integer cosignerCreditScore;
@Column(name = "cosigner_income")
private BigDecimal cosignerIncome;
@Column(name = "relationship_type")
private String relationshipType;
protected CosignerSecuredOffer() {}
public CosignerSecuredOffer(MyLoanOffer originalOffer) {
super(originalOffer, MyProcuringTypes.COSIGNER);
}
@Override
public String getOfferKey() {
return getOriginalOffer().getUuid() + ":COSIGNER";
}
// Getters/setters omitted for brevity
}
Creating Procuring Engines
1. Vehicle Collateral Engine
@Component
public class VehicleProcuringEngine implements ProcuringEngine {
@Override
public ProcuringType procuringType() {
return MyProcuringTypes.VEHICLE;
}
@Override
public Collection<? extends SecuredOffer> generateSecuredOffers(ProductOffer productOffer) {
MyLoanOffer offer = (MyLoanOffer) productOffer;
VehicleSecuredOffer securedOffer = new VehicleSecuredOffer(offer);
// Apply vehicle-specific business logic
BigDecimal baseRate = offer.getProductAdditive().getInterestRate();
// Vehicle collateral typically reduces rate by 1-2%
BigDecimal adjustedRate = baseRate.subtract(BigDecimal.valueOf(0.015));
securedOffer.setAdjustedInterestRate(adjustedRate);
// Set loan-to-value ratio (typically 80-90% for vehicles)
securedOffer.setLoanToValueRatio(BigDecimal.valueOf(0.85));
return List.of(securedOffer);
}
}
2. Cosigner Engine
@Component
public class CosignerProcuringEngine implements ProcuringEngine {
@Override
public ProcuringType procuringType() {
return MyProcuringTypes.COSIGNER;
}
@Override
public Collection<? extends SecuredOffer> generateSecuredOffers(ProductOffer productOffer) {
MyLoanOffer offer = (MyLoanOffer) productOffer;
CosignerSecuredOffer securedOffer = new CosignerSecuredOffer(offer);
// Cosigner reduces risk, so lower interest rate
BigDecimal baseRate = offer.getProductAdditive().getInterestRate();
BigDecimal adjustedRate = baseRate.subtract(BigDecimal.valueOf(0.02));
securedOffer.setAdjustedInterestRate(adjustedRate);
return List.of(securedOffer);
}
}
3. Property Collateral Engine
@Component
public class PropertyProcuringEngine implements ProcuringEngine {
private final PropertyValuationService propertyService;
public PropertyProcuringEngine(PropertyValuationService propertyService) {
this.propertyService = propertyService;
}
@Override
public ProcuringType procuringType() {
return MyProcuringTypes.PROPERTY;
}
@Override
public Collection<? extends SecuredOffer> generateSecuredOffers(ProductOffer productOffer) {
MyLoanOffer offer = (MyLoanOffer) productOffer;
// Only generate if participant has property
if (!hasEligibleProperty(offer.getParticipant())) {
return Collections.emptyList();
}
PropertySecuredOffer securedOffer = new PropertySecuredOffer(offer);
// Property collateral gets best rates
BigDecimal baseRate = offer.getProductAdditive().getInterestRate();
BigDecimal adjustedRate = baseRate.multiply(BigDecimal.valueOf(0.7)); // 30% reduction
securedOffer.setAdjustedInterestRate(adjustedRate);
return List.of(securedOffer);
}
private boolean hasEligibleProperty(Participant participant) {
// Business logic to check property ownership
return propertyService.hasVerifiedProperty(participant);
}
}
Custom Data Processing
1. Implement Data Processor
@Component
public class MyDataProcessor extends OfferEngineDataProcessor<UUID, MyEntity> {
@Override
public ExecutionResultType getResultType() {
return new ExecutionResultType("MY_LOAN");
}
@Override
public Collection<Map<String, Object>> mapToData(MyEntity entity) {
// Map entity data for offer generation
Map<String, Object> data = new HashMap<>();
data.put("creditScore", entity.getCreditScore());
data.put("income", entity.getMonthlyIncome());
return List.of(data);
}
}
2. Configure Offer Generation Scripts
Scripts evaluate participant data and return offer parameters:
// Example offer generation script
if (profile.creditScore >= 700) {
offer.interestRate = productAdditive.interestRate * 0.9; // 10% discount
offer.maxAmount = Math.min(profile.income * 12, productAdditive.maxAmount);
return offer;
} else if (profile.creditScore >= 600) {
offer.interestRate = productAdditive.interestRate;
offer.maxAmount = Math.min(profile.income * 8, productAdditive.maxAmount);
return offer;
} else {
return null; // No offer for low credit scores
}
18.8. Next Steps
-
[credit-management]: Convert selected offers into active credits
-
Workflow Integration: Automate offer generation through workflows
-
[operations]: Handle offer-related operations and modifications
-
[payment-transactions]: Process payments for accepted offers
The Offer Engine provides the foundation for personalized lending by transforming static products into dynamic, risk-adjusted offers tailored to each participant’s profile and circumstances.
19. REST API Integration
TimveroOS provides a complete REST API framework that runs as a separate application alongside your main admin interface. This architecture enables clean separation between internal operations and external integrations.
19.1. Multi-Application Architecture
TimveroOS applications can run multiple Spring Boot contexts simultaneously, each serving different purposes:
public class ExampleApplication {
public static void main(String[] args) {
SpringApplicationBuilder parentBuilder = new SpringApplicationBuilder(BaseConfiguration.class, CustomConfiguration.class)
.web(WebApplicationType.NONE);
parentBuilder.run(args);
// Admin interface on port 8081
parentBuilder.child(WebMvcConfig.class, CustomWebConfiguration.class)
.properties("server.port=8081")
.run(args);
// REST API on port 8082
parentBuilder.child(PortalWebConfiguration.class)
.properties("server.port=8082")
.run(args);
}
}
This pattern allows you to:
-
Scale independently - API and admin interface can have different resource requirements
-
Secure separately - Different authentication mechanisms for different audiences
-
Deploy flexibly - API can be deployed to different environments or behind different load balancers
-
Version independently - API versioning without affecting admin functionality
19.2. API Application Setup
The API application requires minimal configuration:
@SpringBootApplication
@EnableAutoConfiguration
@ComponentScan(basePackageClasses = {ApiWebConfig.class, PortalWebConfiguration.class})
public class PortalWebConfiguration {
}
Key components provided by the framework:
-
ApiWebConfig- Configures REST endpoints, serialization, and validation -
OpenAPI Integration - Automatic Swagger documentation generation
-
Exception Handling - Standardized error responses
-
Request/Response Processing - JSON serialization with proper HTTP status codes
19.3. Controller Patterns
TimveroOS API controllers follow enterprise patterns with comprehensive documentation:
@RestController
@RequestMapping("clients")
@SecurityRequirement(name = BASIC_AUTH)
@Tag(name = "Client Management", description = "API for managing clients")
public class ClientController {
@PostMapping
@Operation(summary = "Create a new client")
@ApiResponses(value = {
@ApiResponse(responseCode = "200", description = "Client created successfully"),
@ApiResponse(responseCode = "400", description = "Invalid request data")
})
public ResponseEntity<CreateClientResponse> createClient(
@RequestBody @Valid CreateClientRequest form) {
UUID clientId = clientService.createClient(form);
return ResponseEntity.ok(new CreateClientResponse(clientId));
}
}
Framework Advantages:
-
Automatic validation - Jakarta Bean Validation with detailed error messages
-
OpenAPI generation - Complete API documentation without manual maintenance
-
Type safety - Request/response DTOs with MapStruct mapping
-
Consistent error handling - Framework-provided exception responses
19.4. Service Layer Integration
API controllers leverage your existing business logic without duplication:
@Service
public class ClientService {
@Autowired
private ClientRequiestMapper mapper; // MapStruct generated
@Autowired
private ClientRepository clientRepository; // Your existing repository
@Transactional
public UUID createClient(@Valid CreateClientRequest form) {
Client client = mapper.createEntity(form); // DTO to Entity mapping
Client savedClient = clientRepository.save(client);
return savedClient.getId();
}
}
This approach ensures:
-
Single source of truth - Business logic remains in service layer
-
Consistent validation - Same rules for API and admin interface
-
Transaction management - Proper database transaction handling
-
Audit trails - All operations logged through existing mechanisms
19.5. Authentication & Security
The framework provides configurable authentication:
@Configuration
@EnableWebSecurity
@SecurityScheme(type = SecuritySchemeType.HTTP, name = "basicAuth", scheme = "basic")
public class InternalApiSecurityConfig {
@Value("${internal.api.username:username}")
private String internalApiUser;
@Value("${internal.api.password:password}")
private String internalApiPassword;
}
Security Features:
-
Basic Authentication - Simple, reliable for internal APIs
-
Configurable credentials - Environment-specific configuration
-
Swagger UI protection - Documentation access control
-
Method-level security -
@PreAuthorizesupport for fine-grained access
Alternative Authentication Methods
The framework supports multiple authentication strategies beyond Basic Auth:
JWT Token Authentication:
@SecurityScheme(
type = SecuritySchemeType.HTTP,
name = "bearerAuth",
scheme = "bearer",
bearerFormat = "JWT"
)
@Configuration
public class JwtSecurityConfig {
@Bean
public SecurityFilterChain jwtFilterChain(HttpSecurity http) throws Exception {
http.oauth2ResourceServer(oauth2 -> oauth2.jwt(Customizer.withDefaults()));
return http.build();
}
}
OAuth2 Integration:
@SecurityScheme(
type = SecuritySchemeType.OAUTH2,
name = "oauth2",
flows = @OAuthFlows(
authorizationCode = @OAuthFlow(
authorizationUrl = "${oauth2.authorization-uri}",
tokenUrl = "${oauth2.token-uri}",
scopes = {
@OAuthScope(name = "read", description = "Read access"),
@OAuthScope(name = "write", description = "Write access")
}
)
)
)
API Key Authentication:
@SecurityScheme(
type = SecuritySchemeType.APIKEY,
name = "apiKey",
in = SecuritySchemeIn.HEADER,
paramName = "X-API-Key"
)
public class ApiKeySecurityConfig {
@Bean
public SecurityFilterChain apiKeyFilterChain(HttpSecurity http) throws Exception {
http.addFilterBefore(new ApiKeyAuthenticationFilter(), UsernamePasswordAuthenticationFilter.class);
return http.build();
}
}
Multiple Authentication Schemes: You can configure multiple authentication methods simultaneously:
@RestController
@SecurityRequirements({
@SecurityRequirement(name = "basicAuth"),
@SecurityRequirement(name = "bearerAuth")
})
public class FlexibleAuthController {
// Accepts both Basic Auth and JWT tokens
}
Authentication Strategy Selection:
-
Basic Auth - Internal tools, development environments
-
JWT/OAuth2 - External integrations, mobile applications
-
API Keys - Third-party service integrations, webhooks
-
Mutual TLS - High-security B2B integrations
-
Custom schemes - Proprietary authentication systems
19.6. Request/Response Mapping
MapStruct integration provides type-safe, performant mapping:
@Mapper
public interface ClientRequiestMapper {
@Mapping(target = "individualInfo.nationalId", source = "nationalId")
@Mapping(target = "contactInfo.email", source = "email")
@Mapping(target = "participants", ignore = true)
Client createEntity(CreateClientRequest form);
}
Mapping Advantages:
-
Compile-time generation - No runtime reflection overhead
-
Type safety - Compilation errors for mismatched fields
-
Nested object support - Complex object graph mapping
-
Validation integration - Works seamlessly with Jakarta Bean Validation
19.7. API Documentation
OpenAPI documentation is generated automatically and includes:
-
Interactive Swagger UI - Available at
/swagger-ui.html -
Complete schema definitions - Request/response models with validation rules
-
Authentication flows - Security requirements clearly documented
-
Example requests - Generated from your validation annotations
Access the documentation at: http://localhost:8082/swagger-ui.html
19.8. Advanced Capabilities
Webhook Support:
@RestController
@RequestMapping("webhooks")
public class WebhookController {
@PostMapping("/docusign")
public ResponseEntity<Void> handleDocusignWebhook(@RequestBody DocusignEvent event) {
// Process external webhook
return ResponseEntity.ok().build();
}
}
Async Processing:
@PostMapping("/applications")
public ResponseEntity<CreateApplicationResponse> createApplication(@RequestBody CreateApplicationRequest request) {
UUID applicationId = applicationService.createApplicationAsync(request);
return ResponseEntity.accepted().body(new CreateApplicationResponse(applicationId));
}
File Upload Support:
@PostMapping(value = "/documents", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
public ResponseEntity<UploadResponse> uploadDocument(@RequestParam("file") MultipartFile file) {
// Framework handles file processing
return ResponseEntity.ok(uploadResponse);
}
19.9. Integration Patterns
External System Integration: The API layer serves as an integration point for external systems while maintaining all business rules and validation from your main application.
Mobile Application Backend: Provides clean JSON APIs for mobile applications with proper error handling and validation.
Third-party Service Integration: Webhooks and callbacks from external services (payment processors, document signing, credit bureaus) can be handled through dedicated API endpoints.
Microservices Communication: When scaling to microservices architecture, the API layer provides a stable interface contract between services.
For detailed Spring Boot and OpenAPI configuration, refer to:
20. Docusign Integration
This section describes how to integrate Docusign electronic signature service into your Timvero application for document signing workflows.
20.1. Integration Overview
The framework reduces DocuSign integration to three simple steps:
-
Entity implements DocusignSigner - delegate to existing entity fields
-
Service calls framework - one method call handles everything
-
Optional webhook - extend default for custom business logic
What the framework handles: - DocuSign API authentication and session management - Envelope creation and document upload - Signer management and embedded URL generation - Status polling fallback when webhooks fail - Multi-party signing workflows with routing order - Document template integration for signature placement
What you implement: - Business logic for document selection based on application state - Mapping entity fields to signer information - Custom webhook logic (optional)
20.2. Configuration Setup
Properties Configuration
Add the following properties to your application configuration file (e.g., application.properties, application.yml, or environment-specific property files):
# DocuSign Configuration
docusign.api.base.path=https://demo.docusign.net/restapi
docusign.oauth.base.path=account-d.docusign.com
docusign.rsa.key.file=classpath:/docusign/private.key
docusign.client.id=602c4320-5b82-48bf-b0c9-9d47a84de053
docusign.user.id=c2e5a6c9-fd35-477f-8104-e2e5385d251a
| The configuration values shown above are examples only. Your actual configuration will differ and should be obtained from your DocuSign account and environment setup. |
Property Descriptions
docusign.api.base.path-
Base URL for DocuSign REST API endpoints. The actual URLs may vary depending on your DocuSign environment and region. Refer to the DocuSign platform documentation for current API endpoints.
docusign.oauth.base.path-
OAuth base domain for DocuSign JWT authentication. Check the DocuSign platform documentation for the appropriate OAuth endpoints for your environment.
docusign.rsa.key.file-
Spring Resource path to the RSA private key file used for JWT authentication. Supports prefixes like
classpath:,file:, etc. docusign.client.id-
Integration Key (Client ID) for your DocuSign application. This value can be obtained from your DocuSign account on the platform.
docusign.user.id-
DocuSign User ID (GUID) associated with your integration. This value can be obtained from your DocuSign account on the platform.
| All DocuSign-specific configuration values (API endpoints, Client ID, User ID) should be obtained from your DocuSign account dashboard. Refer to the DocuSign developer documentation for detailed instructions on locating these values. |
Private Key Setup
The DocuSign integration requires an RSA private key for JWT authentication. The docusign.rsa.key.file property uses Spring Resource syntax:
Classpath resource:
docusign.rsa.key.file=classpath:/docusign/private.key
Place file at: src/main/resources/docusign/private.key
File system resource:
docusign.rsa.key.file=file:/path/to/docusign/private.key
Use absolute path to file on server filesystem.
| For detailed instructions on generating and configuring DocuSign API credentials and private keys, refer to the DocuSign JWT Authentication Guide. |
20.3. Implementation Steps
Step 1: Entity Implementation
Your entity must implement the DocusignSigner interface to provide signer information:
@Entity
public class Participant extends AbstractAuditable implements DocusignSigner {
@Override
public String getSignerName() {
return getClient().getIndividualInfo().getFullName();
}
@Override
public String getSignerEmail() {
return getClient().getContactInfo().getEmail();
}
// For multi-party signing (optional)
@Override
public List<DocusignSignerInfo> getAllSigners() {
// Default: single signer with routing order 1
return List.of(new DocusignSignerInfo(getId(), getSignerName(), getSignerEmail(), 1, "borrower"));
}
}
Key points:
- Delegate to existing entity fields - no new data storage needed
- getAllSigners() enables multi-party workflows and document template integration
- Default implementation handles single signer scenarios
Step 2: Service Integration
Implement your business logic and call the framework:
@Service
public class ApplicationPortalService {
@Autowired
private DocusignSignatureService docusignSignatureService;
@Transactional
public String getSignatureUrl(UUID applicationId, String returnUrl) throws IOException, SignatureException {
// Your business logic
Application application = findApplication(applicationId);
SignableDocument document = selectDocumentBasedOnStatus(application);
Participant participant = application.getBorrowerParticipant();
// Framework handles everything else
return docusignSignatureService.getDocusignUrl(participant, document, returnUrl);
}
}
Framework call does: - Creates DocuSign envelope if needed - Uploads document to DocuSign - Manages signer creation - Returns embedded signing URL - Handles status polling fallback
Status-based document selection pattern:
SignableDocumentType documentType;
switch (portalStatus) {
case IN_PROCESS -> documentType = APPLICATION_FORM;
case PENDING_CONTRACT_SIGNATURE -> documentType = APPLICATION_CONTRACT;
case null, default -> throw new PreconditionFailedException(
"Signature is not available for application status: " + portalStatus);
}
Step 3: Controller Endpoint
Standard REST controller delegates to service:
@GetMapping("/signature-url")
@Operation(summary = "Get application signature url")
@ApiResponses(value = {
@ApiResponse(responseCode = "200", description = "Signature URL retrieved successfully"),
@ApiResponse(responseCode = "404", description = "Application not found"),
@ApiResponse(responseCode = "412", description = "Application with incorrect status provided")
})
public ResponseEntity<String> getApplicationSignatureUrl(
@RequestParam UUID applicationId,
@RequestParam String returnUrl) throws IOException, SignatureException {
String signatureUrl = applicationService.getSignatureUrl(applicationId, returnUrl);
return ResponseEntity.ok(signatureUrl);
}
Step 4: Webhook Implementation (Optional)
The framework provides automatic webhook handling at /callback/docusign/webhook. When documents are signed, the framework automatically processes them without additional code.
Webhook resilience: Framework ignores webhooks for unknown envelopes and polls DocuSign directly when webhooks fail.
Custom Webhook (Optional)
Extend the default webhook for additional business logic:
@RestController
@RequestMapping(value = DocusignWebhookController.PATH, produces = MediaType.APPLICATION_JSON_VALUE)
public class DocusignWebhookController {
@Autowired
private DocusignSignatureService docusignSignatureService;
@PostMapping(value = "/webhook", produces = "application/json;charset=UTF-8")
public void handleWebhook(@RequestBody DocusignWebhookResponse payload) throws IOException {
if (payload.getData() != null && payload.getData().getUserId().equals(userId)) {
String envelopeId = payload.getData().getEnvelopeId();
try {
docusignSignatureService.signDocumentByEnvelopeId(envelopeId);
// Add your custom logic here: notifications, status updates, etc.
} catch (SignatureException e) {
// Framework ignores unknown signatures - webhook may arrive for
// signatures not tracked in our system
}
}
}
}
20.4. Document Templates (Responsive Signing)
DocuSign Responsive Signing allows creating HTML documents with embedded signing fields that automatically adapt to mobile devices. The framework integrates with getAllSigners() to map template roles to actual signers.
Important: Responsive Signing must be enabled by DocuSign support. Contact DocuSign to request "Enable Smart Sections/API for Responsive Signing" for your account.
DocuSign HTML Elements
Signature and identity fields:
<!-- Signature field -->
<ds-signature data-ds-role="borrower"></ds-signature>
<!-- Full name (auto-populated) -->
<ds-fullname data-ds-role="borrower"></ds-fullname>
<!-- Date signed (auto-populated) -->
<ds-date-signed data-ds-role="borrower"></ds-date-signed>
Input fields using standard HTML with DocuSign attributes:
<!-- Text input -->
<input data-ds-type="text" data-ds-role="borrower" />
<!-- Email input with validation -->
<input data-ds-type="email" data-ds-role="borrower" />
<!-- Optional field -->
<input data-ds-type="text" data-ds-role="borrower" required="false" />
Field Requirements
By default, all fields are required. To make a field optional:
<input data-ds-type="text" data-ds-role="borrower" required="false" />
Styling and Customization
All standard HTML styling is supported:
<!-- Styled signature -->
<ds-signature data-ds-role="borrower"
style="width:300px;height:100px;border:2px solid #000;">
</ds-signature>
<!-- Styled date field -->
<ds-date-signed data-ds-role="borrower"
style="color:red;font-size:18px;font-weight:bold;">
</ds-date-signed>
<!-- Text input with styling -->
<input data-ds-type="text"
data-ds-role="borrower"
style="width:200px;font-family:Arial;"
id="borrowerPhone"
class="form-field" />
Complete Template Example
<!DOCTYPE html>
<html>
<head>
<title>Loan Application</title>
</head>
<body>
<h1>Personal Loan Agreement</h1>
<h3>Borrower Information:</h3>
<p>Email: <input data-ds-type="email" data-ds-role="borrower" /></p>
<p>Name: <ds-fullname data-ds-role="borrower"></ds-fullname></p>
<p>Phone: <input data-ds-type="text" data-ds-role="borrower" required="false" /></p>
<div>
<br><br>
Borrower Signature: <ds-signature data-ds-role="borrower"></ds-signature>
</div>
<p>Date Signed: <ds-date-signed data-ds-role="borrower"></ds-date-signed></p>
</body>
</html>
Framework Integration
Template roles must match timveroRole values from getAllSigners():
@Override
public List<DocusignSignerInfo> getAllSigners() {
return List.of(
new DocusignSignerInfo(borrowerId, borrowerName, borrowerEmail, 1, "borrower"),
new DocusignSignerInfo(guarantorId, guarantorName, guarantorEmail, 2, "guarantor")
);
}
<!-- Template uses matching role names -->
<ds-signature data-ds-role="borrower"></ds-signature>
<ds-signature data-ds-role="guarantor"></ds-signature>
Form Data Collection
Input fields automatically populate the formData map in DocusignDocumentSignature:
<input data-ds-type="text" data-ds-role="borrower" name="income" />
<input data-ds-type="email" data-ds-role="borrower" name="contact_email" />
After signing, access the data:
DocusignDocumentSignature signature = (DocusignDocumentSignature) document.getSignature();
Map<String, String> formData = signature.getFormData();
String income = formData.get("income");
String email = formData.get("contact_email");
Key requirements:
-
data-ds-rolevalues must exactly matchtimveroRolefromgetAllSigners() -
Responsive Signing must be enabled for your DocuSign account
-
All fields are required by default unless
required="false"is specified -
Standard HTML styling and attributes are fully supported
Additional Resources
For more detailed information about DocuSign Responsive Signing and HTML templates:
Official DocuSign Documentation:
-
How To Create a Signable HTML Document - Main guide for creating HTML signing documents
-
Responsive Signing Concepts - Core concepts and principles
DocuSign Developer Blog:
-
From the Trenches: Please sign responsively - Detailed guide with code examples for
<ds-date-signed>,<ds-signature>and other tags -
Use the Apex Toolkit to send responsive HTML documents - Examples using
<input data-ds-type="email">,<ds-fullname>and<ds-signature>
Code Examples:
-
DocuSign Template Library - HTML Examples - Real-world HTML template with all field types
DocuSign Webhook Configuration
Configure webhooks in your DocuSign account:
-
Go to your DocuSign Admin panel
-
Navigate to Integrations > Webhooks
-
Create a new webhook with your endpoint URL
-
Select relevant events (e.g., "Envelope Completed")
For detailed configuration instructions, see the DocuSign Webhooks Documentation.
21. Covenant Monitoring System
The Covenant Monitoring System provides automated control of the financial conditions of loan agreements through integration with the Metric System and the Feature Store.
21.1. Architecture Overview
The Covenant Monitoring System follows a four-layer architecture:
-
CovenantSpecification — versioned monitoring rules
-
CovenantExecution — scheduling and performing checks
-
CovenantResult — storing check results and managing violations
-
Metric Integration — calculation of indicators via
MetricService
Scheduled Task → CovenantSpecification → MetricService → CovenantExecution → CovenantResult
↓
Feature Store + DataSources
Key Principle: Covenants always operate through the Metric System.
Indicators are computed via MetricMapping, and validation conditions are defined through Expression.
Core System Components
| Component | Description |
|---|---|
CovenantSpecification |
A versioned configuration of a monitoring rule. Defines the holder, subject, metrics, and validation conditions. |
CovenantExecution |
A record of an executed check. Contains status, a collection of results, and error information. |
CovenantResult |
The outcome of a check for a specific subject. Possible states: |
HasCovenant |
Marker interface for entities that own covenants (e.g., |
HasMetric |
Marker interface for entities being evaluated (e.g., |
CovenantHolderResolver |
Links a holder with its subjects and defines the logic for resolving the entities to be validated. |
What the Framework Does
Automatic Monitoring:
-
Periodic Execution — checks are performed on a schedule
-
Metric Calculation — via Feature Store and DataSources
-
CovenantExecution Creation — generated for each scheduled check
-
Result Persistence — stores outcomes for all subjects
Lifecycle Management:
-
Producer — creates tasks with status
CovenantExecution.NEW -
Consumer — executes the checks and saves the results
-
Exception Handling — automatic processing of metrics in
EXCEPTIONstate -
Versioning — managed through
lineageId
What is a Covenant?
A Covenant is a contractual financial condition that a borrower agrees to maintain throughout the term of a loan. Covenants are used for risk monitoring and early detection of financial distress.
Typical Covenant Examples:
-
Borrower Income — must not decrease by more than 20% from the baseline value
-
Loan-to-Value (LTV) — loan amount to collateral value ratio < 80%
-
Debt-to-Income (DTI) — debt-to-income ratio < 40%
-
Minimum Liquidity — available funds > €10,000
CovenantSpecification — Rule Configuration
CovenantSpecification defines what to check, who to check, and how often the checks are performed.
| Field | Description |
|---|---|
holderType |
The type of entity that owns the covenant ( |
subjectKey |
Specifies who is being checked: |
metricMapping |
The |
anchoredMetricMapping |
Optional |
expression |
JavaScript/Groovy expression used to evaluate the condition.
Returns a |
executionType |
Execution type. Currently only |
periodicity |
Frequency of execution: |
numberOfPeriods |
Number of periods between executions (e.g., |
dateTimeExecution |
Date and time of the first scheduled execution. The Producer calculates subsequent ones based on periodicity. |
additives |
List of product additive IDs. The covenant applies only to loans with these products. |
active |
Activity flag. Only active specifications are executed. |
Versioning:
-
CovenantSpecificationextendsHistoryEntity— versioning is supported -
All versions are linked through a shared
lineageId -
When a specification is edited, a new version is created with
active=true -
Previous versions remain in history with
active=false
CovenantExecution — Check Record
CovenantExecution is created for each run of a check based on a CovenantSpecification.
| Field | Description |
|---|---|
specification |
Reference to the |
periodIndex |
The index of the period for scheduled executions. Calculated based on |
status |
|
covenantResults |
A collection of |
exception |
|
createdAt |
Timestamp when the execution was created. |
createdBy |
User who triggered the execution (for manual runs) or |
CovenantResult — Check Result
CovenantResult stores the outcome of a check for a specific subject within a CovenantExecution.
| Field | Description |
|---|---|
execution |
Reference to the corresponding |
specification |
Reference to the |
ownerId |
UUID of the holder entity (e.g., |
subjectId |
UUID of the evaluated subject (e.g., |
metric |
Reference to the computed |
rawContent |
The raw result of the expression evaluation (string representation). |
state |
|
createdAt |
Timestamp of result creation. |
CovenantResult States:
-
{ok} CLEAN — condition satisfied, no violations
-
{bad} VIOLATION — condition violated
-
{ok} RESOLVED — violation manually resolved by a user
-
{bad} EXCEPTION — error occurred during metric or expression evaluation
Metric Types: REGULAR vs ANCHORED
The system uses two types of metrics for different monitoring purposes.
REGULAR Metrics
Purpose: Represents the current value of an indicator during each evaluation.
When calculated: On every covenant execution via MetricService.getMetric().
Examples:
-
Current monthly income of the borrower
-
Current market value of the collateral
-
Current debt-to-income ratio
ANCHORED Metrics
Purpose: Fixed baseline value used for comparison.
When set:
-
At loan origination — automatically via
AnchoredMetricService.calculateAnchored() -
When conditions change — recalculated via
recalculateAnchored() -
Manually by a user — through
UpdateAnchoredMetricAction
Examples:
-
Income at the time of loan origination
-
Initial appraised value of the collateral
-
DTI ratio at loan approval
HasCovenant Interface
public interface HasCovenant extends Persistable<UUID> {
}
Purpose: Marks entities that can have covenants (holders).
Examples:
-
ExampleCredit— loans with financial conditions
Required to implement:
-
CovenantHolderResolver<T extends HasCovenant>for this type -
Define subjects via
getCovenantSubjects()
HasMetric Interface
public interface HasMetric extends FeaturedSubject {
default String details() {
return getPrimaryId();
}
}
Purpose: Marks entities that can be evaluated (subjects).
Examples:
-
Participant— borrowers and guarantors -
Collateral— pledged assets
Required to implement:
-
details()— human-readable description for the UI (name, ID, etc.) -
getPrimaryId()— unique identifier for logs
CovenantHolderResolver — Linking holder ↔ subjects
CovenantHolderResolver defines relationships between a holder and its subjects.
| Method | Description |
|---|---|
resolveTargets |
|
getCovenantSubjects |
|
getCovenantSpecifications |
|
21.2. Complete Implementation Example
Below is a complete implementation of the covenant system for a credit product using ExampleCredit as the holder.
Overview
In this example, we implement borrower income monitoring to verify that income has not dropped by more than 20% from the baseline captured at loan origination.
Implementation components:
-
CovenantHolderResolver — links
ExampleCreditwithParticipantsubjects -
MetricMapping — config for computing income via the Feature Store
-
CovenantSpecification — the income verification rule
-
UI Integration — tabs to browse results
Step 1: CovenantHolderResolver Implementation
Create a resolver that links the credit to the participants being evaluated.
@Component
public class CreditCovenantHolderResolver implements CovenantHolderResolver<ExampleCredit> {
@Autowired
private ExampleCreditRepository creditRepository;
@Autowired
private CovenantSpecificationRepository covenantSpecificationRepository;
@Override
public Stream<ExampleCredit> resolveTargets(CovenantSpecification specification) {
// Find all credits to check by product additive
return specification.getAdditives().stream()
.flatMap(a -> creditRepository.getAllByAdditiveId(a));
}
@Override
public List<SubjectRecord<ExampleCredit>> getCovenantSubjects() {
return List.of(
// BORROWER subject
SubjectRecordBuilder.<ExampleCredit>builder()
.name(ParticipantRole.BORROWER.name())
.subject(new SubjectSupplierRecord<>(
Participant.class,
credit -> List.of(credit.getApplication().getBorrowerParticipant())
))
.build(),
// GUARANTOR subject
SubjectRecordBuilder.<ExampleCredit>builder()
.name(ParticipantRole.GUARANTOR.name())
.subject(new SubjectSupplierRecord<>(
Participant.class,
credit -> credit.getApplication().getParticipants().stream()
.filter(p -> p.getRoles().contains(ParticipantRole.GUARANTOR))
.collect(Collectors.toList())
))
.build()
);
}
@Override
public Collection<CovenantSpecification> getCovenantSpecifications(ExampleCredit target) {
// Find all active specs for the product additive of this credit
return covenantSpecificationRepository.findAllByAdditiveIdAndActiveTrue(
target.getCondition().getSecuredOffer().getOriginalOffer().getProductAdditive().getId()
);
}
}
Key Points:
-
@Component — Spring automatically discovers the resolver via its generic type
-
resolveTargets — finds credits by the product additive from the specification
-
getCovenantSubjects — three subject types: BORROWER, GUARANTOR, COLLATERAL
-
getCovenantSpecifications — filters specs by the credit’s product additive
Step 2: Entity Integration
Ensure that entities implement the required interfaces.
ExampleCredit Entity
@Entity
@DiscriminatorValue("1")
public class ExampleCredit extends Credit implements HasCovenant {
@OneToOne(fetch = FetchType.LAZY, optional = false)
@JoinColumn(unique = true, nullable = false)
private BaseApplication application;
// ... other fields
}
Participant Entity
@Entity
@Table(name = "participant")
@Audited
public class Participant extends AbstractAuditable
implements HasMetric {
@Override
public String details() {
return getFullName(); // "John Smith"
}
@Transient
@Override
public String getPrimaryId() {
IndividualInfo info = getClient() != null
? getClient().getIndividualInfo()
: getIndividualInfo();
return info.getNationalId(); // "123456789"
}
// ... other fields
}
Important:
-
HasCovenant — marker for holder entities
-
HasMetric — marker for subject entities
-
details() — human-readable name for the UI
-
getPrimaryId() — unique ID for logs and grouping
Step 3: MetricMapping Configuration
Create a MetricMapping in the UI to compute a participant’s monthly income.
Configuration:
| Field | Value |
|---|---|
Name |
|
Entity Type |
|
Expression |
JavaScript code that extracts income |
Engine |
|
Expression for computing income:
// Extract income from the Feature Store
var incomeData = features.get("tink", "monthly_income");
// If no banking data, fall back to manually provided values
if (incomeData == null) {
var annualIncome = entity.getTotalAnnualIncome();
if (annualIncome != null) {
return annualIncome.getNumber().doubleValue() / 12;
}
return null;
}
// Return monthly income from banking data
return incomeData;
What happens:
-
features.get() — loads data from the Feature Store (e.g., Tink bank data)
-
Fallback — if no external data is available, use
entity.getTotalAnnualIncome() -
Automatic typing —
MetricServiceinfers theFeatureDataTypefrom the result
Step 4: CovenantSpecification Configuration
Create a CovenantSpecification in the UI to validate income.
Core parameters:
| Field | Value |
|---|---|
Name |
|
Holder Type |
|
Subject Key |
|
Metric Mapping |
|
Anchored Metric Mapping |
|
Execution Type |
|
Periodicity |
|
Number of Periods |
|
Date Time Execution |
|
Additives |
Select the product additives to which the rule applies |
Validation expression:
// Ensure income has not dropped by more than 20%
// metric — current income (REGULAR)
// anchored — baseline income at origination (ANCHORED)
if (anchored == null) {
// If no baseline is set, skip the check
return true;
}
// Income must be >= 80% of the baseline
return metric >= anchored * 0.8;
Bindings available in the expression:
| Binding | Description |
|---|---|
|
|
|
|
|
|
|
Direct access to the metric value ( |
|
Value of the ANCHORED metric or |
Step 5: ANCHORED Metric Setup
Set the baseline income value at the time of loan origination.
Automatic Setup
When the credit transitions to an active state, invoke:
@Component
public class CreditChecker extends EntityChecker<ExampleCredit, UUID> {
@Autowired
private AnchoredMetricService anchoredMetricService;
@Override
protected void registerListeners(CheckerListenerRegistry<ExampleCredit> registry) {
registry.entityChange().inserted();
}
@Override
protected boolean isAvailable(ExampleCredit credit) {
return true;
}
@Override
protected void perform(ExampleCredit credit) {
TransactionUtils.afterTransaction(() -> {
anchoredMetricService.calculateAnchored(credit);
});
}
}
What happens:
-
Finds all
CovenantSpecificationswithanchoredMetricMapping != null -
Resolves subjects via
CovenantHolderResolver -
Computes metrics via
MetricService -
Persists them as
MetricType.ANCHORED
Manual Setup via UI
A user can update an ANCHORED metric with UpdateAnchoredMetricAction:
@RequestMapping("/update-anchored-metric")
@Controller
@Order(10000)
public class UpdateAnchoredMetricAction extends AbstractUpdateAnchoredMetricAction<ExampleCredit> {
// Inherits full functionality
}
UI flow:
-
Open the credit → Actions → Update Anchored Metric
-
Select the subject (borrower)
-
Select the metric (
participant_monthly_income) -
Enter a new value and data type
-
Save — a new ANCHORED metric is created
Step 6: UI Integration
Add tabs to display check results.
Tab with the Latest Results
@RequestMapping("/covenant-result")
@Controller
@Order(9500)
public class CreditCovenantResultTab extends AbstractCovenantResultTab<ExampleCredit> {
@Autowired
private CovenantResultRepository covenantResultRepository;
@Override
protected List<CovenantResult> getCovenantResults(UUID id) {
// Shows only the latest result for each subject
return covenantResultRepository.findLatestByOwnerGroupedBySubjectId(id);
}
}
Tab with Full History
@RequestMapping("/covenant-result-archive")
@Controller
@Order(10500)
public class CreditCovenantResultArchiveTab extends AbstractCovenantResultTab<ExampleCredit> {
@Autowired
private CovenantResultRepository covenantResultRepository;
@Override
protected List<CovenantResult> getCovenantResults(UUID id) {
// Show all check results
return covenantResultRepository.findAllByOwnerId(id);
}
}
Tab for ANCHORED Metrics
@RequestMapping("/anchored-metric")
@Controller
@Order(10000)
public class CreditAnchoredMetricTab extends HasCovenantAnchoredMetricTab<ExampleCredit> {
@Override
public boolean isVisible(ExampleCredit entity) {
return true; // Always visible
}
}
Action to calculate the covenant manually:
@RequestMapping("/calculate-covenant")
@Controller
@Order(11000)
public class CalculateCovenantAction extends AbstractCalculateCovenantAction<ExampleCredit> {
// Inherits full functionality
}
Localization
Add translations for UI elements in messages.properties.
#---------------Actions-------------
credit.action.update-anchored-metric=Update Anchored Metric
credit.action.calculate-covenant=Calculate Covenant
#---------------Tabs----------------
credit.tab.covenant-result=Covenant Results
credit.tab.covenant-result-archive=Covenant Results Archive
credit.tab.anchored-metric=Anchored Metrics
Step 7: Testing
Validate the implementation via the UI.
Test Scenario 1: Creating a Specification
-
Go to
/covenant-specification→ Create New -
Fill in all fields (see Step 4)
-
Use the Test button with a real credit
-
Verify the expression returns
true/false -
Save the specification
Test Scenario 2: Setting the ANCHORED Metric
-
Open a credit in status NEW
-
Activate the credit →
calculateAnchored()is invoked -
Navigate to the Anchored Metrics tab
-
Verify that income is stored with type ANCHORED
21.3. Expression Writing Guide
An expression defines the covenant condition. It is JavaScript or Groovy code that must return a boolean result.
Expression Basics
Expression requirements:
-
Returns boolean —
trueforCLEAN,falseforVIOLATION -
Short and readable — avoid overly complex logic
-
Null-safe — always check for
nullbefore use -
Deterministic — same input → same output
Supported scripting engines:
-
JavaScript — recommended; fast and simple
-
Groovy — for advanced logic with Java interop
Available Bindings
An expression has access to the following variables:
| Binding | Type | Description |
|---|---|---|
|
|
The entity that owns the covenant. All fields accessible via getters. |
|
|
The entity being evaluated. Fields accessible via getters. |
|
|
Metric object with methods |
|
Depends on the metric |
Direct access to the metric value. The name matches |
|
Depends on the metric or |
Value of the ANCHORED metric. |
Expression Examples
Example 1: Debt-to-Income Check
Requirement: DTI < 40%.
// Get the monthly payment from the credit schedule
var monthlyPayment = covenantHolder.getSchedule().getPaymentAmount().getNumber().doubleValue();
// Get the current monthly income from the metric
var monthlyIncome = metric;
if (monthlyIncome == null || monthlyIncome <= 0) {
return false; // No income = VIOLATION
}
// DTI must be < 40%
var dti = monthlyPayment / monthlyIncome;
return dti < 0.4;
Example 2: Combined Check
Requirement: Income remains stable AND DTI is acceptable.
// Check 1: income has not dropped by more than 20%
var incomeStable = anchored != null && metric >= anchored * 0.8;
// Check 2: DTI < 40%
var monthlyPayment = covenantHolder.getSchedule().getPaymentAmount().getNumber().doubleValue();
var dtiAcceptable = (monthlyPayment / metric) < 0.4;
// Both conditions must pass
return incomeStable && dtiAcceptable;
Testing Expressions
Use the built-in expression tester in the UI for debugging.
Performance Considerations
Each expression is executed for every subject during every evaluation.
Anti-Patterns
-
{bad} Not checking null — causes
EXCEPTIONinstead ofVIOLATION -
{bad} Overly nested logic — makes expressions unreadable
-
{bad} Iterating over collections — poor performance impact
-
{bad} DB queries inside expression — use only
MetricMapping+Feature Store -
{bad} Mutation inside expression — expressions must only evaluate, not modify state
-
{bad} Using console.log — not available in production, only for debugging
-
{bad} Hardcoded entity IDs — expressions must work for any entity
-
{bad} Ignoring currency — always validate the currency for
MonetaryAmount
Expression Checklist
Before saving a CovenantSpecification, verify the following:
-
{todo} Expression returns a boolean
-
{todo} All null cases are handled
-
{todo} Tested with real data via UI
-
{todo} Works for all subject types (BORROWER, GUARANTOR, etc.)
-
{todo} Handles the case when
anchored = null -
{todo} No magic numbers — use clear constants
-
{todo} Comments added for non-trivial logic
-
{todo} Under 20 lines of code
21.4. Lifecycle & Scheduling
The Covenant Monitoring System uses a producer–consumer pattern to automatically execute checks on a schedule.
Overview
Execution Flow:
CovenantSpecification (active=true)
↓
Producer Task (scheduled) → creates CovenantExecution.NEW
↓
Consumer Task (scheduled) → processes NEW → saves Results
↓
CovenantResult (CLEAN / VIOLATION / EXCEPTION)
Key Components:
-
CovenantExecutionProducerTask — scheduled task that creates executions
-
CovenantExecutionProducer — business logic responsible for creation
-
CovenantExecutionConsumer — scheduled task that processes executions
-
CovenantExecutionService — performs checks and persists results
Producer Task
The Producer creates CovenantExecution records with status NEW for further processing.
Configuration
# application.properties
covenant.execution.producer.initialDelay=PT1M
covenant.execution.producer.fixedDelay=PT5M
Properties:
| Property | Description |
|---|---|
|
Delay before the first execution after application startup. Format: ISO-8601 Duration ( |
|
Interval between the completion of the previous run and the start of the next one. Recommended: |
Key Points:
-
Idempotent — repeated runs do not create duplicates (checked via
exists) -
Period-based — only one execution per period
-
REQUIRES_NEW — independent transaction for each specification
-
Synchronous load —
getSync()ensures consistency
Consumer Task
The Consumer processes executions with status NEW and saves the results.
Configuration
# application.properties
covenant.execution.consumer.initialDelay=PT2M
covenant.execution.consumer.fixedDelay=PT1M
Properties:
| Property | Description |
|---|---|
|
Delay before the first run. Should be greater than the Producer’s delay to allow time for execution creation. |
|
Interval between runs. Recommended: |
Processing Strategy:
-
One at a time — processes one execution at a time
-
FIFO order —
ORDER BY id DESC LIMIT 1(can be changed to ASC) -
Separate transaction — each execution runs in its own transaction
-
Fail-safe — failure in one execution does not block others
21.5. Best Practices
Guidelines for designing, implementing, and operating the Covenant Monitoring System.
CovenantSpecification Design
Naming Conventions
-
Descriptive names — use “Borrower Income Stability > 80%” instead of “Income Check”
-
Include thresholds — explicitly state numeric thresholds in names
-
Subject prefix — start with “Borrower”, “Guarantor”, or “Collateral”
-
Version suffix — add “v2”, “Q1 2025”, etc., for multiple versions
Examples:
Good examples:
-
Borrower Monthly Income > 80% of Baseline
-
Guarantor Active Status Check
-
Collateral LTV < 80%
-
Borrower DTI < 40% (Revised Q1 2025)
Bad examples:
-
{bad} Inc Check
-
{bad} GUAR_STS
-
{bad} LTV
Expression Design
-
Keep it simple — under 20 lines, one main logical condition
-
Null-safe first — check for null at the beginning
-
Use constants — avoid magic numbers
-
Comment thresholds — explain where threshold values come from
-
Return explicit boolean — always use
return true/false -
Handle missing anchored — define behavior when the baseline is null
-
Test all edge cases — null, zero, and negative values
Good example:
// Borrower monthly income must not drop below 80% of baseline
// Threshold agreed with Risk Management on 2025-01-15
if (anchored == null) {
return true; // Skip check if baseline not set
}
if (metric == null || metric <= 0) {
return false; // No income data = violation
}
var THRESHOLD = 0.8; // 20% drop allowed
return metric >= anchored * THRESHOLD;
Bad example:
// Don’t do this
return metric >= anchored * 0.8; // What if anchored is null?
CovenantHolderResolver Implementation
Resolver Design
-
Single responsibility — one resolver per holder type
-
Immutable subjects list — use
List.of()forgetCovenantSubjects() -
Efficient resolveTargets — use indexed queries
-
Filter inactive — exclude cancelled or closed entities
-
Handle lazy loading — use
JOIN FETCHwhen needed
Good example:
@Override
public Stream<ExampleCredit> resolveTargets(CovenantSpecification specification) {
// Efficient query with filtering
return specification.getAdditives().stream()
.flatMap(additiveId -> creditRepository
.findByAdditiveIdAndStatusIn(
additiveId,
List.of(CreditStatus.ACTIVE, CreditStatus.GRACE_PERIOD)
)
.stream()
);
}
Bad example:
@Override
public Stream<ExampleCredit> resolveTargets(CovenantSpecification specification) {
// Inefficient — loads all credits
return creditRepository.findAll().stream()
.filter(c -> specification.getAdditives().contains(c.getAdditive().getId()));
}
21.6. Anti-Patterns
Common mistakes when working with the Covenant Monitoring System and how to avoid them.
Expression Anti-Patterns
Not Checking Null Values
Problem: The expression throws a NullPointerException instead of returning VIOLATION.
Bad example:
// Dangerous — if anchored == null, this will throw an exception
return metric >= anchored * 0.8;
Why it’s bad:
-
{bad} Creates a
CovenantResultwithstate = EXCEPTION -
{bad} Masks the real violation
-
{bad} Makes problem analysis harder
-
{bad} Requires manual intervention
Good example:
// Safe — explicit null check
if (anchored == null) {
return true; // Or false, depending on business logic
}
if (metric == null) {
return false; // No data = violation
}
return metric >= anchored * 0.8;
CovenantHolderResolver Anti-Patterns
Loading All Entities
Problem: The resolver loads all entities without any filtering.
Bad example:
@Override
public Stream<ExampleCredit> resolveTargets(CovenantSpecification specification) {
// Loads ALL credits from the database
return creditRepository.findAll().stream()
.filter(c -> c.getStatus() == CreditStatus.ACTIVE);
}
Why it’s bad:
-
{bad} Huge memory footprint
-
{bad} Very slow on large datasets
-
{bad} Execution timeouts
-
{bad} Risk of OutOfMemoryError when scaling
Good example:
@Override
public Stream<ExampleCredit> resolveTargets(CovenantSpecification specification) {
// Efficient query with filtering at the DB level
return specification.getAdditives().stream()
.flatMap(additiveId -> creditRepository
.findByAdditiveIdAndStatus(additiveId, CreditStatus.ACTIVE)
.stream()
);
}
Using REGULAR Instead of ANCHORED
Problem: Attempting to compare two REGULAR metrics instead of REGULAR vs ANCHORED.
Bad example:
CovenantSpecification: Metric Mapping: participant_monthly_income (REGULAR) Anchored Metric Mapping: null Expression: // Tries to compare with a previous value // But there is no anchored metric! return metric >= previousValue * 0.8; // previousValue is unknown
Why it’s bad:
-
{bad} No baseline for comparison
-
{bad} Expression has no access to history
-
{bad}
anchored = null→ check is skipped
Good example:
CovenantSpecification:
Metric Mapping: participant_monthly_income (REGULAR)
Anchored Metric Mapping: participant_monthly_income (ANCHORED)
Expression:
if (anchored == null) {
return true;
}
return metric >= anchored * 0.8;
21.7. Production Checklist
Final checklist before launching the Covenant Monitoring System in production.
CovenantSpecification Configuration
-
{todo} Specification has a clear, descriptive name
-
{todo} Subject key matches the keys defined in the resolver
-
{todo} Metric mapping exists and is active
-
{todo} Expression tested with real data via the UI tester
-
{todo} Expression handles all null cases
-
{todo} Periodicity matches the frequency of data updates
-
{todo} DateTimeExecution is set correctly
-
{todo} Product additives selected properly
-
{todo} Only one active version (unless A/B testing)
-
{todo} Expression < 20 lines of code
-
{todo} Threshold values documented (e.g., source of 0.8, 0.4, etc.)
CovenantHolderResolver Implementation
-
{todo} Resolver is registered as a Spring @Component
-
{todo} resolveTargets() uses indexed queries
-
{todo} resolveTargets() filters only active/relevant entities
-
{todo} getCovenantSubjects() returns an immutable List
-
{todo} Subject suppliers return List (not null) for empty collections
-
{todo} Subject suppliers use eager fetch or JOIN FETCH
-
{todo} getCovenantSpecifications() filters by active=true
-
{todo} No N+1 query problems during subject extraction
-
{todo} Resolver covered by unit tests
MetricMapping Configuration
-
{todo} MetricMapping has a clear, descriptive name
-
{todo} Entity type matches the subject class
-
{todo} Expression returns a consistent type
-
{todo} Expression handles DataUnavailableException
-
{todo} Fallback chain defined (external → manual → null)
-
{todo} Expression tested through the UI
-
{todo} Feature Store integration configured
-
{todo} DataSource timeout properly set
-
{todo} MetricMapping versioned if modified
-
{todo} Active = true for all used mappings
Expression Validation
-
{todo} Checks for null before using anchored
-
{todo} Checks for null before using metric
-
{todo} Returns explicit boolean (true/false)
-
{todo} No magic numbers present
-
{todo} Uses constants with meaningful names
-
{todo} Comments explain threshold values
-
{todo} Tested for edge cases (null, zero, negative)
-
{todo} Works for all subject types
ANCHORED Metrics Setup
-
{todo} ANCHORED metrics set during disbursement/activation
-
{todo} calculateAnchored() called at the correct lifecycle event
-
{todo} Manual update via UI works correctly
-
{todo} Update requires approval (role-based access)
-
{todo} No automated scheduled updates
-
{todo} ANCHORED MetricMapping exists for all related specifications
22. Credit Management System
This section describes how to implement and manage credit entities using the loan servicing framework. The ExampleCredit implementation demonstrates a complete loan management system, but the underlying loan module provides flexible components for implementing various lending products.
22.1. Credit System Architecture
The credit management system is built on the loan servicing framework, which provides core components for any type of lending product. The ExampleCredit serves as a reference implementation, but you can create different credit types for various lending scenarios.
Core Components
The loan module (com.timvero.servicing) provides the foundation:
-
Credit- Base entity class for all lending products -
CreditSnapshot- Point-in-time credit state management -
CreditOperation- Base class for all credit operations (payments, charges, etc.) -
Debt- Flexible debt structure with account-based balances -
CreditCalculationService- Core calculation engine -
CreditPaymentService- Payment processing infrastructure
ExampleCredit Implementation
The ExampleCredit demonstrates a complete consumer loan implementation:
@Entity
@DiscriminatorValue("1")
public class ExampleCredit extends Credit implements NamedEntity, HasCovenant {
@NotNull
@OneToOne(fetch = FetchType.EAGER)
@JoinColumn(nullable = false)
private Application application;
@OneToOne(fetch = FetchType.LAZY, cascade = CascadeType.ALL)
@JoinColumn(name = "condition")
@Fetch(FetchMode.JOIN)
private ExampleCreditCondition condition;
public Application getApplication() {
return application;
}
public void setApplication(Application application) {
this.application = application;
}
public ExampleCreditCondition getCondition() {
return condition;
}
public void setCondition(ExampleCreditCondition condition) {
this.condition = condition;
}
@Override
public String getDisplayedName() {
return "Loan for " + getApplication().getBorrowerParticipant().getDisplayedName();
}
@Transient
public LocalDate getMaturityDate() {
return getStartDate().plus(getCondition().getPeriod().multipliedBy(getCondition().getTerm()));
Key features:
-
Discriminator Value:
"1"identifies this credit type in the database -
Application Integration: Links to loan application and borrower
-
Condition Management: Contains loan terms (principal, term, interest rate)
-
Maturity Calculation: Automatic calculation based on start date and terms
22.2. Credit Entity Setup
Creating Custom Credit Types
To implement different lending products, extend the base Credit class:
@Entity
@DiscriminatorValue("2") // Unique identifier for this credit type
public class MortgageCredit extends Credit implements NamedEntity {
@OneToOne(fetch = FetchType.EAGER)
@JoinColumn(nullable = false)
private PropertyApplication application;
@OneToOne(fetch = FetchType.LAZY, cascade = CascadeType.ALL)
@JoinColumn(name = "mortgage_condition")
private MortgageCondition condition;
@Override
public String getDisplayedName() {
return "Mortgage for " + getApplication().getPropertyAddress();
}
@Transient
public LocalDate getMaturityDate() {
return getStartDate().plus(getCondition().getTerm());
}
}
Credit Condition Configuration
Define loan terms and conditions specific to your credit type:
@Entity
public class ExampleCreditCondition extends BasePersistable<UUID> {
@Embedded
@CompositeType(MonetaryAmountType.class)
private MonetaryAmount principal;
@Column(nullable = false)
private Period period; // Payment frequency (monthly, weekly, etc.)
@Column(nullable = false)
private Integer term; // Number of payment periods
@Column(nullable = false)
private BigDecimal interestRate;
// getters and setters...
}
Database Schema Generation
The platform automatically generates SQL migrations for credit entities:
-- Generated migration for ExampleCredit
CREATE TABLE credit (
id UUID PRIMARY KEY,
credit_type INTEGER NOT NULL, -- Discriminator column
start_date DATE NOT NULL,
calculation_date DATE,
actual_snapshot BIGINT,
created_at TIMESTAMP NOT NULL,
updated_at TIMESTAMP NOT NULL,
-- ExampleCredit specific columns
application UUID NOT NULL,
condition UUID
);
22.3. Credit Operations Framework
Built-in Operation Types
The loan module provides standard operation types that work with any credit implementation:
Charge Operations
Implement fees and charges specific to your credit type:
@Entity
@DiscriminatorValue("301")
public class OriginationFeeCharge extends CreditOperation {
public static Integer TYPE = 301;
@Embedded
@CompositeType(MonetaryAmountType.class)
private MonetaryAmount amount;
public OriginationFeeCharge(LocalDate date, MonetaryAmount amount) {
super(TYPE, date, OperationStatus.APPROVED);
this.amount = amount;
}
@Override
public boolean isEndDayOperation() {
return false;
}
}
Accrual Operations
Interest and fee accruals are handled by the calculation engine:
@Entity
@DiscriminatorValue("401")
public class InterestAccrual extends CreditOperation {
@Embedded
@CompositeType(MonetaryAmountType.class)
private MonetaryAmount accruedAmount;
@Column(nullable = false)
private String accountType; // INTEREST, LATE_FEE, etc.
@Override
public boolean isEndDayOperation() {
return true; // Accruals typically run at end of day
}
}
Credit Action Controllers
Actions provide user interface operations for credit management:
Payment Registration
@Controller
@RequestMapping("/register-payment")
public class RegisterPaymentAction extends EntityActionController<UUID, ExampleCredit, ManualTransferForm> {
@Autowired
private BorrowerTransactionService borrowerTransactionService;
public static final Long OTHER = 0L;
@Override
protected EntityAction<? super ExampleCredit, ManualTransferForm> action() {
return when(c -> c.getActualSnapshot() != null && c.getActualSnapshot().getStatus().equals(ACTIVE))
.then((c, f, u) -> {
LiquidityClientPaymentMethod paymentMethod =
new LiquidityClientPaymentMethod(f.getProcessedDate(), f.getAmount(), TransactionType.INCOMING,
c.getApplication().getBorrowerParticipant().getClient().getIndividualInfo().getFullName());
borrowerTransactionService.proceedCustom(c, TransactionType.INCOMING, paymentMethod,
paymentMethod.getAmount(), true, f.getDescription());
});
The action uses a form to collect payment details:
public static class ManualTransferForm {
@NotNull
@Positive
private MonetaryAmount amount;
@NotNull
@DateTimeFormat(pattern = PATTERN_DATEPICKER_FORMAT)
private LocalDate processedDate;
private String description;
public MonetaryAmount getAmount() {
return amount;
}
public void setAmount(MonetaryAmount amount) {
this.amount = amount;
}
public LocalDate getProcessedDate() {
return processedDate;
}
public void setProcessedDate(LocalDate processedDate) {
this.processedDate = processedDate;
}
public String getDescription() {
return description;
}
public void setDescription(String description) {
this.description = description;
}
}
Disbursement Processing
@Controller
@RequestMapping("/register-disbursement")
public class RegisterDisbursementAction extends EntityActionController<UUID, ExampleCredit, DisbursementForm> {
@Override
protected EntityAction<? super ExampleCredit, DisbursementForm> action() {
return when(c -> c.getActualSnapshot().getStatus().equals(APPROVED))
.then((c, f, u) -> {
// Create disbursement transaction
disbursementService.processDisbursement(c, f.getAmount(), f.getMethod());
});
}
}
Transaction Processing Integration
The credit system integrates with the transaction processing framework:
@Autowired
private PaymentTransactionService transactionService;
@Autowired
private CoreCreditRepository creditRepository;
@Autowired
private PaymentTransactionRepository transactionRepository;
@Autowired
private BorrowerTransactionRepository borrowerTransactionRepository;
@Autowired
private CreditPaymentService paymentService;
@Autowired
private ChargeOperationService chargeOperationService;
@Transactional(propagation = Propagation.MANDATORY)
public void proceedCustom(ExampleCredit credit, TransactionType type, PaymentMethod paymentMethod,
MonetaryAmount amount, boolean sync, String description) {
BorrowerTransaction transaction = new BorrowerTransaction(type, amount, paymentMethod, credit);
transaction.setStatus(TransactionStatus.READY_FOR_EXECUTION);
transaction.setDescription(description);
Transaction processing flow:
-
Transaction Creation:
BorrowerTransactioncreated with payment details -
Payment Gateway: Transaction sent to payment processor
-
Success Handling: On success, creates
CreditPaymentoperation -
Calculation Trigger: Credit calculation engine updates balances
-
Snapshot Update: New credit state snapshot created
22.4. Credit Calculation Engine
Debt Structure
The loan module uses a flexible debt structure with named accounts:
// Account type constants for ExampleCredit
public static final String PRINCIPAL = "PRINCIPAL";
public static final String INTEREST = "INTEREST";
public static final String LATE_FEE = "LATE_FEE";
public static final String PAST_DUE_PRINCIPAL = "PAST_DUE_PRINCIPAL";
public static final String PAST_DUE_INTEREST = "PAST_DUE_INTEREST";
For different credit types, define appropriate account structures:
// Mortgage-specific accounts
public static final String PRINCIPAL = "PRINCIPAL";
public static final String INTEREST = "INTEREST";
public static final String ESCROW = "ESCROW";
public static final String PMI = "PMI";
public static final String PROPERTY_TAX = "PROPERTY_TAX";
Calculation Service Integration
The calculation engine automatically processes credit operations:
// Triggered after payment registration
creditCalculationService.calculate(creditId, paymentDate, currentCalculationDate);
This recalculates:
-
Debt Balances: Updates account balances based on payment distribution
-
Interest Accruals: Calculates daily interest charges
-
Past Due Amounts: Moves overdue balances to past due accounts
-
Credit Status: Updates credit status based on payment history
22.5. Credit User Interface
Controller Implementation
@RequestMapping(value = ExampleCreditController.PATH)
@MenuItem(order = 5_300, name = "credit")
public class ExampleCreditController extends ViewableFilterController<UUID, ExampleCredit, ExampleCreditFilter> {
public static final String PATH = "/credit";
@Override
protected String getHeaderPage() {
return "/credit/header";
The controller provides:
-
List View: Paginated credit listing with filtering
-
Detail View: Comprehensive credit information display
-
Action Buttons: Context-sensitive operations based on credit status
Credit Filtering
public class ExampleCreditFilter extends ListFilter {
@Field(restriction = Restriction.IN, value = ExampleCredit_.ACTUAL_SNAPSHOT + "." + CreditSnapshot_.STATUS)
private CreditStatus[] status;
public CreditStatus[] getStatus() {
return status;
}
Filtering capabilities:
-
Status Filtering: Filter by credit status (ACTIVE, CLOSED, etc.)
-
Date Ranges: Filter by creation date, maturity date, etc.
-
Amount Ranges: Filter by principal amount, current balance
-
Custom Filters: Add business-specific filtering criteria
Tab-Based Interface
Credit details are organized into tabs for better user experience:
// Credit data tab showing balances and payment history
@Controller
@Order(1000)
public class CreditDataTab extends EntityTabController<UUID, ExampleCredit> {
@Override
protected String getTabTemplate(UUID id, Model model) throws Exception {
ExampleCredit credit = loadEntity(id);
// Add credit summary data
model.addAttribute("currentBalance", credit.getActualSnapshot().getDebt());
model.addAttribute("paymentHistory", getPaymentHistory(credit));
model.addAttribute("nextPaymentDate", calculateNextPaymentDate(credit));
return super.getTabTemplate(id, model);
}
}
Available tabs:
-
Credit Data: Current balances, payment status, maturity information
-
Payments: Payment history and upcoming payment schedule
-
Transactions: All financial transactions related to the credit
-
Calculations: Detailed calculation history and interest computations
-
Documents: Credit-related documents and contracts
Form Components for Credit Operations
Credit forms use the platform’s form component system:
<!-- Payment amount input -->
<th:block th:insert="~{/form/components :: amount(
#{credit.payment.amount},
'amount',
'v-required v-positive')}"
th:with="currencies = ${currencies}" />
<!-- Payment date picker -->
<th:block th:insert="~{/form/components :: date(
#{credit.payment.date},
'processedDate',
'v-required')}"
th:with="minDate = ${minDate}" />
<!-- Payment description -->
<th:block th:insert="~{/form/components :: textarea(
#{credit.payment.description},
'description',
'')}"
th:with="rows = 3" />
22.6. Credit Business Rules with Entity Checkers
Automated Credit Monitoring
Entity checkers can implement automated credit monitoring:
@Component
public class CreditPaymentChecker extends EntityChecker<ExampleCredit> {
@Override
protected void registerListeners(CheckerListenerRegistry<ExampleCredit> registry) {
// Monitor payment operations
registry.entityChange(CreditPayment.class, payment ->
creditRepository.findByOperationsIn(payment))
.inserted();
}
@Override
protected boolean isAvailable(ExampleCredit credit) {
return credit.getActualSnapshot().getStatus().equals(ACTIVE);
}
@Override
protected void perform(ExampleCredit credit) {
// Check if credit is now current after payment
if (isPaidCurrent(credit)) {
updateCreditStatus(credit, CURRENT);
sendPaymentConfirmation(credit);
}
}
}
Past Due Processing
Automated past due detection and processing:
@Component
public class PastDueChecker extends EntityChecker<ExampleCredit> {
@Override
protected void registerListeners(CheckerListenerRegistry<ExampleCredit> registry) {
// Monitor daily calculation updates
registry.entityChange().updated("calculationDate");
}
@Override
protected boolean isAvailable(ExampleCredit credit) {
return isPastDue(credit) && !isAlreadyMarkedPastDue(credit);
}
@Override
protected void perform(ExampleCredit credit) {
// Create past due operation
PastDueOperation pastDue = new PastDueOperation(
LocalDate.now(),
calculatePastDueAmount(credit)
);
credit.getOperations().add(pastDue);
// Send past due notification
notificationService.sendPastDueNotice(credit);
}
}
22.7. Extending the Credit System
Creating New Credit Types
To implement different lending products:
-
Extend Credit Entity: Create new entity with specific discriminator value
-
Define Condition Structure: Create condition entity with product-specific terms
-
Implement Operations: Create custom payment and charge operation types
-
Configure Account Types: Define debt account structure for the product
-
Create Controllers: Implement UI controllers and actions
-
Add Business Rules: Implement entity checkers for automated processing
Integration with External Systems
The credit system supports integration with external services:
-
Payment Gateways: Transaction processing through payment providers
-
Credit Bureaus: Credit reporting and monitoring
-
Core Banking: Integration with banking systems
-
Document Management: Contract and document storage
-
Notification Services: Email and SMS notifications
22.8. Best Practices
Credit Design Principles
-
Separation of Concerns: Keep business logic in services, not entities
-
Event-Driven Architecture: Use entity checkers for automated processing
-
Flexible Debt Structure: Design account types to accommodate future requirements
-
Audit Trail: Leverage built-in auditing for compliance requirements
-
Transaction Safety: Ensure operations are atomic and consistent
Performance Considerations
-
Calculation Optimization: Batch credit calculations during off-peak hours
-
Snapshot Management: Archive old snapshots to maintain performance
-
Index Strategy: Create appropriate database indexes for credit queries
-
Lazy Loading: Use lazy loading for large collections and related entities
Security and Compliance
-
Data Protection: Implement field-level encryption for sensitive data
-
Access Control: Use role-based security for credit operations
-
Audit Logging: Maintain comprehensive audit trails for regulatory compliance
-
Data Retention: Implement data retention policies for closed credits
The credit management system provides a robust foundation for implementing various lending products while maintaining consistency, auditability, and extensibility across different credit types.
23. Credit Operations Framework
The Credit Operations Framework provides a powerful, extensible system for managing financial operations throughout the credit lifecycle. This framework handles everything from loan disbursements and payments to interest accruals and account closures, with full audit trails and automated processing capabilities.
|
The |
23.1. Operations Architecture Overview
The operations framework is built on a flexible, event-driven architecture that separates operation definitions from their processing logic, enabling customization for various lending products and business models.
Core Components
The loan servicing module (com.timvero.servicing) provides the foundation:
-
CreditOperation- Base entity for all credit operations -
CreditOperationHandler<O>- Interface for operation processing logic -
Snapshot- Point-in-time credit state representation -
AccrualEngine- Interface for time-based calculations -
PreCalculateSynchronizer- Interface for operation synchronization -
CreditCalculationService- Core calculation engine that processes operations
23.2. Operation Processing Flow
The credit calculation system processes operations through a sophisticated pipeline that ensures correct chronological execution and state management.
The Calculation Pipeline
When CreditCalculationService.calculate() is called, the system follows this flow:
-
Synchronization Phase:
PreCalculateSynchronizerimplementations (likeAccrualOperationService) ensure all necessary operations exist -
Date Range Processing: The system processes each date from the start date to the target date
-
Daily Operation Processing: For each date, operations are sorted by their
getOrder()value and processed sequentially -
Snapshot Creation: After processing all operations for a date, a
CreditSnapshotis created and stored -
State Updates: The credit’s actual snapshot and calculation date are updated
Operation Processing Order
Operations are processed in a specific order determined by their getOrder() method. Lower numbers execute first:
| Order | Operation Type | Example Type Code | Purpose |
|---|---|---|---|
101 |
Charge Operations |
901 |
Add fees, penalties, or other charges to the credit |
111 |
Accrual Operations |
950 |
Calculate and apply interest, late fees, and other time-based charges |
200 |
Payment Operations |
200 |
Process borrower payments and apply to debt balances |
900 |
Past Due Operations |
900 |
Move current debt to past due status when payments are missed |
995 |
Void Operations |
995 |
Cancel or void credit operations |
999 |
Close Operations |
999 |
Close and finalize credit accounts |
End-of-Day vs Intraday Operations
Each operation implements isEndDayOperation() to control when during the day it should be processed:
-
isEndDayOperation() = false: Intraday operations - processed immediately when encountered -
isEndDayOperation() = true: End-of-day operations - processed only when the date is "closed"
This distinction is crucial for business logic:
Intraday Operations (Most Operations)
@Override
public boolean isEndDayOperation() {
return false; // Process immediately
}
-
Charges - Applied immediately when created
-
Payments - Processed as soon as received
-
Accruals - Applied when calculation runs
End-of-Day Operations (Special Cases)
@Override
public boolean isEndDayOperation() {
return true; // Process only at end of day
}
-
Past Due Operations - Only processed when the day is "closed" to ensure all payments for that day have been received
This design prevents premature past due processing if a payment arrives later in the same business day.
Operation Handler Execution
Within each date, the OperationProcessor handles individual operations:
-
Handler Discovery: The system finds the appropriate
CreditOperationHandlerfor each operation type -
Snapshot Application: Each handler modifies the current
Snapshotto reflect the operation’s effect -
Debt Tracking: The system tracks how each operation changes the debt balances
-
State Management: Operations can modify both debt balances and credit status
Transaction and Locking
The calculation service uses sophisticated transaction management:
-
Pessimistic Locking: Credits are locked during calculation to prevent concurrent modifications
-
Separate Transactions: Synchronization and calculation run in separate transactions
-
Event Publishing: Status changes and snapshot updates trigger events for other system components
23.3. Operation Entity Structure
Base Operation Entity
All credit operations extend the base CreditOperation class:
@Entity
@Inheritance(strategy = InheritanceType.SINGLE_TABLE)
@DiscriminatorColumn(name = "operation_type", discriminatorType = DiscriminatorType.INTEGER)
public abstract class CreditOperation extends AbstractAuditable<UUID> {
@Column(nullable = false)
private Integer type;
@Column(nullable = false)
private LocalDate date;
@Enumerated(EnumType.STRING)
@Column(nullable = false)
private OperationStatus status;
// Abstract methods that subclasses must implement
public abstract boolean isEndDayOperation();
public abstract int getOrder();
}
Key features:
-
Single Table Inheritance: All operations stored in one table with discriminator
-
Type Identification: Each operation type has a unique integer identifier
-
Date-based Processing: Operations are tied to specific business dates
-
Status Management: Operations can be APPROVED, CANCELED, DECLINED, or PENDING
-
Audit Trail: Full change history via
AbstractAuditable
23.4. Implementing Custom Operations
To implement custom operations, follow the same patterns demonstrated in the example project. You’ll need to:
-
Create the Operation Entity - Extend
CreditOperationwith your specific fields and behavior -
Implement the Operation Handler - Create a service that implements
CreditOperationHandler<YourOperation> -
Configure the Handler - Add the handler as a Spring bean in your configuration
-
Define Processing Order - Set appropriate
getOrder()andisEndDayOperation()values
The example project’s ChargeOperation and ChargeOperationService demonstrate the simplest implementation pattern, while AccrualOperation and AccrualOperationService show more complex synchronization behavior.
23.5. Standard Operation Types
The example project demonstrates key operation types that represent the core concepts of credit operations.
Charge Operations
Charge operations represent the simplest operation type - they add a monetary amount to a credit account.
@Entity
@DiscriminatorValue("901")
@Audited(targetAuditMode = RelationTargetAuditMode.NOT_AUDITED)
public class ChargeOperation extends CreditOperation {
public static Integer TYPE = 901;
@Embedded
@NotNull
private MonetaryAmount amount;
protected ChargeOperation() {
super();
}
public ChargeOperation(LocalDate date, MonetaryAmount amount) {
super(TYPE, date, OperationStatus.APPROVED);
this.amount = amount;
}
public MonetaryAmount getAmount() {
return amount;
}
@Override
public boolean isEndDayOperation() {
return false;
}
@Override
public int getOrder() {
return 101;
}
}
Key characteristics:
-
Type Code:
901- Unique identifier for database discrimination -
Order:
101- Early processing order to apply charges before other operations -
Monetary Amount: Embedded amount to be added to the debt
Payment Operations
Payment operations process borrower payments and distribute them across debt accounts according to business rules.
The payment distribution order is configured in CreditCalculationConfiguration:
@Bean
CreditPaymentOperationHandler creditPaymentOperationHandler() {
return new CreditPaymentOperationHandler(OVERPAYMENT, List.of(PAST_DUE_PRINCIPAL, PAST_DUE_INTEREST,
LATE_FEE, INTEREST, PRINCIPAL)) {};
}
This defines the payment waterfall: past due amounts first, then fees, current interest, and finally principal. Overpayments are credited to the OVERPAYMENT account.
Accrual Operations
Accrual operations represent a sophisticated concept for calculating time-based charges like interest and late fees. Unlike other operations that are created manually, accrual operations are automatically synchronized with significant credit events.
The Accrual Concept
The key insight behind accrual operations is that interest and fees need to be calculated whenever the debt balance changes. The system automatically creates accrual operations on dates when:
-
Payments are made - Interest must be calculated up to the payment date
-
Past due events occur - Balances move to past due status, affecting future calculations
This synchronization is handled by the AccrualOperationService which implements PreCalculateSynchronizer. It scans for payment and past due operations and ensures corresponding accrual operations exist for those dates.
Accrual Engines
The actual calculation logic is delegated to specialized accrual engines:
-
InterestAccrualEngine- Calculates interest on thePRINCIPALbalance using the credit’s interest rate -
LateFeeAccrualEngine- Calculates late fees on past due principal and interest using the late fee rate
These engines extend BasisAccrualEngine which provides sophisticated day-count calculations, handling rate changes over time, and pro-rated calculations between significant dates.
How It Works
-
Event Detection: When payments or past due operations are processed, the synchronizer identifies dates needing accrual calculations
-
Accrual Creation:
AccrualOperationentities are automatically created for these dates -
Engine Calculation: During credit calculation, accrual engines compute the exact amounts based on outstanding balances, rates, and time periods
-
Balance Application: The calculated accruals are added to the appropriate debt accounts (
INTEREST,LATE_FEE, etc.)
This design ensures that time-based charges are accurately calculated and applied, even when payments are made on irregular dates or when credit terms change over time.
Past Due Operations
Past due operations handle one of the most critical business processes in lending - managing overdue debt. When borrowers miss scheduled payments, the system must reorganize debt balances to reflect the new risk profile and enable different treatment of overdue amounts.
The Past Due Concept
The fundamental principle is that overdue debt behaves differently from current debt:
-
Late fees accrue only on past due balances, not current balances
-
Collection processes target past due amounts with different strategies
-
Reporting and risk assessment treat past due debt as higher risk
-
Payment distribution prioritizes past due amounts over current debt
When a scheduled payment date passes without sufficient payment, current debt must be "moved" to past due accounts to enable this differentiated treatment.
Scheduled Payment Detection
Past due operations are triggered by the credit’s payment schedule, which is defined in the credit condition. The system monitors for:
-
Regular payment dates - monthly, weekly, or other periodic payments based on the credit terms
-
Maturity date - the final payment date when the entire remaining balance becomes due
-
Missed payment amounts - comparing expected payments against actual payments received
Debt Movement Logic
When a past due event occurs, the operation performs account transfers:
-
INTERESTbalance →PAST_DUE_INTERESTaccount -
PRINCIPALbalance →PAST_DUE_PRINCIPALaccount
This reorganization enables the accrual engines to calculate late fees specifically on past due amounts, while new interest continues to accrue on any remaining current principal.
Maturity vs Regular Past Due
The maturity flag in the operation distinguishes between two scenarios:
-
Regular Past Due: Borrower missed a scheduled payment but loan hasn’t matured - partial amounts may move to past due
-
Maturity Past Due: Loan has reached its final payment date - typically the entire remaining balance becomes past due
This distinction allows for different business rules, such as accelerated collection procedures or different late fee calculations for matured loans.
Integration with Credit Lifecycle
Past due operations integrate with other parts of the system:
-
Accrual Operations: Automatically created to calculate late fees on the newly past due amounts
-
Credit Labels: Display indicators like
PastDueLabelto show past due status without changing the core credit status -
Notification Systems: Often trigger automated borrower communications
-
Collection Workflows: May initiate collection processes or escalation procedures
This design ensures that the transition from current to past due debt is handled consistently and triggers all necessary downstream processes.
23.6. Account Structure and Debt Management
The operations framework uses a flexible account-based debt structure defined in the configuration.
Account Type Constants
The example project defines these account types in CreditCalculationConfiguration:
public static final String PRINCIPAL = "PRINCIPAL";
public static final String INTEREST = "INTEREST";
public static final String PAST_DUE_PRINCIPAL = "PD_PRINCIPAL";
public static final String PAST_DUE_INTEREST = "PD_INTEREST";
public static final String LATE_FEE = "LATE_FEE";
public static final String OVERPAYMENT = "OVERPAYMENT";
These accounts represent different types of debt:
* PRINCIPAL - Outstanding loan principal amount
* INTEREST - Accrued interest charges
* PAST_DUE_PRINCIPAL - Overdue principal amounts
* PAST_DUE_INTEREST - Overdue interest amounts
* LATE_FEE - Late payment penalties
* OVERPAYMENT - Credit balance from overpayments
23.7. Operation Configuration
The operations framework requires careful configuration to define how operations behave, how payments are distributed, and how the overall credit system operates. The example project demonstrates a complete configuration approach.
Configuration Architecture
All operation-related configuration is centralized in CreditCalculationConfiguration, which serves as the single source of truth for:
-
Credit Status Definitions - Available credit states and their properties
-
Account Structure - Debt account types and their relationships
-
Operation Handlers - Services that process each operation type
-
Payment Distribution - Rules for how payments are applied to debt
-
Accrual Engines - Time-based calculation components
-
Credit View Options - UI display configuration
Credit Status Configuration
The framework defines credit statuses with specific properties:
public static final CreditStatus PENDING = new CreditStatus("PENDING", 1000, false);
public static final CreditStatus ACTIVE = new CreditStatus("ACTIVE", 1100, false);
public static final CreditStatus CLOSED = new CreditStatus("CLOSED", 2000, true);
public static final CreditStatus VOID = new CreditStatus("VOID", 2100, true);
Each status includes: * Name: Human-readable identifier * Order: Numeric value for status progression logic * Ending Flag: Whether this status represents a terminal state
Account Structure Configuration
The debt account structure is defined as constants:
public static final String PRINCIPAL = "PRINCIPAL";
public static final String INTEREST = "INTEREST";
public static final String PAST_DUE_PRINCIPAL = "PD_PRINCIPAL";
public static final String PAST_DUE_INTEREST = "PD_INTEREST";
public static final String LATE_FEE = "LATE_FEE";
public static final String OVERPAYMENT = "OVERPAYMENT";
These accounts represent different types of debt and credit balances:
* Current Debt: PRINCIPAL, INTEREST - active loan balances
* Past Due Debt: PAST_DUE_PRINCIPAL, PAST_DUE_INTEREST - overdue amounts
* Penalties: LATE_FEE - fees for late payments
* Credits: OVERPAYMENT - borrower credit balances
Operation Handler Configuration
Each operation type requires a corresponding handler bean:
Charge Operation Handler
@Bean
ChargeOperationService chargeOperationService() {
return new ChargeOperationService();
}
The ChargeOperationService implements CreditOperationHandler<ChargeOperation> and defines how charge operations affect debt balances.
Payment Operation Handler
@Bean
CreditPaymentOperationHandler creditPaymentOperationHandler() {
return new CreditPaymentOperationHandler(OVERPAYMENT, List.of(PAST_DUE_PRINCIPAL, PAST_DUE_INTEREST,
LATE_FEE, INTEREST, PRINCIPAL)) {};
}
The payment handler configuration is critical as it defines:
* Overpayment Account: Where excess payments are credited (OVERPAYMENT)
* Payment Waterfall: The order in which payments are applied to debt accounts
Past Due Operation Handler
@Bean
PastDueOperationService pastDueOperationService() {
LinkedHashMap<String, String> map = new LinkedHashMap<>();
map.put(INTEREST, PAST_DUE_INTEREST);
map.put(PRINCIPAL, PAST_DUE_PRINCIPAL);
return new PastDueOperationService(map);
}
The mapping defines how current debt accounts are transferred to past due accounts when payments are missed.
Accrual Engine Configuration
Accrual engines handle specific types of time-based calculations:
Loan Engine Configuration
The loan engine orchestrates the overall calculation process:
@Bean
LoanEngine loanEngine() {
return new BasicLoanEngine(PENDING);
}
The BasicLoanEngine is initialized with the default credit status (PENDING) for new credits.
Credit View Configuration
The view configuration determines which accounts are displayed in the user interface:
@Bean
CreditViewOptions creditViewOptions() {
return new CreditViewOptions(PRINCIPAL, INTEREST, PAST_DUE_PRINCIPAL, PAST_DUE_INTEREST, LATE_FEE);
}
This configuration excludes the OVERPAYMENT account from standard credit displays, as it represents a credit balance rather than debt.
Customizing Configuration for Different Credit Products
Different credit products require different configurations. Common customization patterns include:
Custom Account Structures and Payment Distribution
// Credit card configuration
public static final String PURCHASES = "PURCHASES";
public static final String CASH_ADVANCES = "CASH_ADVANCES";
public static final String FEES = "FEES";
// Payment order: fees first, then cash advances, then purchases
new CreditPaymentOperationHandler<>("CREDIT_BALANCE",
List.of("FEES", "CASH_ADVANCES", "PURCHASES"));
// Mortgage configuration with escrow
public static final String ESCROW = "ESCROW";
public static final String PMI = "PMI";
// Payment order: fees, past due, escrow, current debt
new CreditPaymentOperationHandler<>("ESCROW_SURPLUS",
List.of("LATE_FEE", "PAST_DUE_PRINCIPAL", "ESCROW", "PRINCIPAL"));
Configuration Validation
The framework automatically validates configuration consistency:
-
Handler Registration: Ensures all operation types have corresponding handlers
-
Account References: Validates that payment distribution references valid account types
-
Engine Registration: Confirms accrual engines are properly registered
-
Status Consistency: Checks that status definitions are logically consistent
Configuration Best Practices
-
Centralized Configuration: Keep all operation configuration in a single class for maintainability
-
Meaningful Constants: Use descriptive names for account types and statuses
-
Documented Relationships: Clearly document how accounts relate to each other
-
Environment-Specific Beans: Use Spring profiles or conditions for product-specific configurations
-
Validation: Implement configuration validation to catch setup errors early
-
Consistent Naming: Use consistent naming patterns across account types and operation handlers
The configuration approach in the example project provides a flexible foundation that can be adapted for various lending products while maintaining consistency and clarity.
23.8. Operation Synchronization
The PreCalculateSynchronizer interface enables operations to maintain consistency with related events.
Synchronization Example
The AccrualOperationService demonstrates how synchronization works by automatically creating accrual operations when payments or past due events occur. The service implements PreCalculateSynchronizer and scans for trigger operations, ensuring that accrual operations exist for dates when debt balances change.
You can examine the complete implementation in the source code: AccrualOperationService.java.
23.9. Real-world Usage Patterns
Understanding how operations work together in practice is essential for implementing robust credit systems. The example project’s test cases demonstrate several real-world scenarios that show the complete operation flow.
Scenario 1: Loan Disbursement and Interest Accrual
This scenario demonstrates the basic credit lifecycle from disbursement through interest calculation.
The Flow
-
Credit Creation: A new credit is created with defined terms (principal amount, interest rate, payment schedule)
-
Principal Disbursement: A
ChargeOperationadds the loan principal to thePRINCIPALaccount -
Time Progression: As time passes, the calculation engine processes each day
-
Interest Accrual:
AccrualOperationServiceautomatically createsAccrualOperationentities for interest calculation -
Balance Updates: Interest is calculated and added to the
INTERESTaccount
Scenario 2: Payment Processing and Distribution
This scenario shows how borrower payments are processed and distributed across debt accounts.
The Flow
-
Outstanding Debt: Credit has balances in multiple accounts (principal, interest, fees)
-
Payment Received: An
ExampleCreditPaymentoperation is created -
Payment Distribution: The payment handler applies the payment according to the configured waterfall
-
Balance Updates: Debt accounts are reduced according to priority order
-
Continued Accruals: Interest continues to accrue on remaining balances
Payment Waterfall Logic
The example project uses this payment priority:
-
PAST_DUE_PRINCIPAL- Overdue principal first -
PAST_DUE_INTEREST- Overdue interest second -
LATE_FEE- Late fees third -
INTEREST- Current interest fourth -
PRINCIPAL- Current principal last
From the Test Cases
The paymentOperation1() and paymentOperation2() tests demonstrate:
Partial Payment Scenario:
Charge Principal → Partial Payment → Verify Interest Paid → Verify Principal Unchanged
Full Payment Scenario:
Charge Principal → Full Payment → Verify Interest Paid → Verify Principal Reduced
Scenario 3: Past Due Processing and Late Fees
This scenario illustrates the complex process of handling missed payments and calculating late fees.
The Flow
-
Payment Due Date: A scheduled payment date arrives
-
Insufficient Payment: Borrower makes no payment or insufficient payment
-
Past Due Operation: System automatically creates
PastDueOperation -
Account Transfers: Current debt moves to past due accounts
-
Late Fee Accrual:
LateFeeAccrualEnginebegins calculating fees on past due amounts -
Payment Priority Change: Future payments prioritize past due amounts
Account Movement Logic
When past due occurs:
-
INTERESTbalance →PAST_DUE_INTERESTaccount -
PRINCIPALbalance →PAST_DUE_PRINCIPALaccount
From the Test Case
The pastDue1() test demonstrates this complex scenario:
Charge Principal → Insufficient Payment → Calculate Past Due Date → Verify Account Transfers → Verify Late Fees
The test shows how the system:
-
Moves unpaid amounts to past due accounts
-
Calculates late fees on the past due balances
-
Maintains accurate balance tracking across account types
Scenario 4: End-of-Day vs Intraday Processing
This scenario highlights the importance of operation timing in business logic.
The Concept
-
Intraday Operations: Charges, payments, and accruals process immediately
-
End-of-Day Operations: Past due operations wait until the business day is "closed"
Business Rationale
Past due operations use isEndDayOperation() = true to prevent premature past due status if a payment arrives later in the same business day.
Example Timeline
9:00 AM - Payment due date arrives
10:00 AM - Borrower payment received (processed immediately)
11:00 AM - Another payment received (processed immediately)
End of Day - Past due operation processes only if still insufficient payment
This prevents false past due situations when payments arrive throughout the day.
Scenario 5: Operation Synchronization in Action
This scenario demonstrates how the synchronization system maintains data consistency.
The Challenge
Accrual operations must exist for every date when: - Payments are made (to calculate interest up to payment date) - Past due events occur (to recalculate accruals on new account structure)
The Solution
AccrualOperationService implements PreCalculateSynchronizer:
-
Scan for Trigger Events: Finds all payment and past due operations
-
Identify Required Dates: Determines which dates need accrual calculations
-
Create Missing Operations: Adds
AccrualOperationentities for missing dates -
Cancel Unnecessary Operations: Removes accrual operations for dates without triggers
Key Patterns from Real Usage
1. Event-Driven Architecture
Operations trigger other operations automatically:
-
Payments trigger accrual calculations
-
Past due events trigger late fee calculations
-
Each operation maintains referential integrity
2. Chronological Processing
The calculation engine processes operations in strict date order:
-
Earlier operations affect later operations
-
Processing order within a date matters (
getOrder()values) -
State changes are cumulative and consistent
Testing Real-World Scenarios
The example project’s CalculationTest demonstrates how to test these scenarios:
-
Setup Realistic Conditions: Create credits with proper terms and schedules
-
Apply Operations in Sequence: Mirror real-world event timing
-
Verify All Effects: Check not just primary changes but secondary effects
-
Test Edge Cases: Include scenarios like overpayments and zero balances
These patterns provide a foundation for implementing robust credit operations that handle the complexity of real-world lending scenarios while maintaining accuracy and auditability.
23.10. Best Practices
Operation Design Principles
-
Immutability: Operations should be immutable once created and approved
-
Idempotency: Operations should produce the same result when applied multiple times
-
Atomicity: Each operation should represent a single, atomic business transaction
-
Auditability: All operations must be fully auditable with complete change history
-
Extensibility: Design operations to be easily extended for new business requirements
Performance Optimization
-
Batch Processing: Group related operations for efficient processing
-
Lazy Loading: Use lazy loading for operation collections to avoid N+1 queries
-
Indexing: Create appropriate database indexes for operation queries
-
Caching: Cache frequently accessed operation handlers and calculation results
Error Handling
-
Validation: Validate operations before processing to catch errors early
-
Retry Logic: Implement retry mechanisms for transient failures
-
Compensation: Design compensation operations for failed transactions
-
Monitoring: Monitor operation processing for performance and error patterns
Security Considerations
-
Authorization: Ensure proper authorization for operation creation and modification
-
Data Protection: Protect sensitive operation data with encryption
-
Audit Logging: Maintain comprehensive audit logs for regulatory compliance
-
Access Control: Implement role-based access control for operation management
23.11. Testing Operations
The operations framework provides comprehensive testing capabilities that allow you to verify operation behavior, calculation accuracy, and integration between different operation types.
Integration Testing Approach
The example project demonstrates a complete integration testing strategy using CalculationTest that tests the entire operation processing pipeline.
Test Configuration
The test uses a comprehensive Spring configuration that mirrors the production setup:
@DataJpaTest
@AutoConfigureEmbeddedDatabase(provider = DatabaseProvider.ZONKY)
@ContextConfiguration(classes = {CreditCalculationConfiguration.class, CalculationTest.CalculationTestConfig.class})
Key testing components:
-
Embedded Database: Uses Zonky for isolated database testing
-
Transaction Management:
TransactionTemplateBuilderfor proper transaction handling -
Real Services: Uses actual
CreditCalculationServiceandAccrualServiceinstances -
Complete Configuration: Includes all operation handlers and accrual engines
Testing Patterns
The example project demonstrates several essential testing patterns:
Testing the complete credit lifecycle from creation through operations:
// Create credit with realistic conditions
UUID creditId = initCredit(startDate, TODAY);
// Apply operations
charge(creditId, chargeDate, principalAmount);
registerPayment(creditId, paymentDate, paymentAmount);
// Trigger calculation
calculate(creditId, startDate, TODAY);
// Verify results
ExampleCredit credit = entityManager.find(ExampleCredit.class, creditId);
Assertions.assertEquals(expectedBalance, credit.getActualSnapshot().getDebt().getAccount(PRINCIPAL).get());
Testing how different operations interact with each other:
@Test
public void paymentOperation1() {
// Setup: Create credit and charge principal
charge(creditId, chargeDate, principal);
// Test: Make partial payment
registerPayment(creditId, paymentDate, partialPayment);
calculate(creditId, startDate, TODAY);
// Verify: Check payment distribution and remaining balances
ExampleCreditPayment payment = credit.getOperations(ExampleCreditPayment.class, APPROVED).findAny().get();
assertEquals(expectedInterestPayment, payment.getFinalDebt().get().getAccount(INTEREST).get());
assertEquals(expectedRemainingInterest, credit.getActualSnapshot().getDebt().getAccount(INTEREST).get());
}
Testing time-based calculations and accrual accuracy:
@Test
public void chargeOperation() {
charge(creditId, chargeDate, principal);
calculate(creditId, startDate, TODAY);
// Verify accrued interest calculation
MonetaryAmount expectedInterest = principal.multiply(
(INTEREST_RATE.doubleValue() / 100d) * (daysBetween / 360d)
);
Debt accruals = accrualService.calculateCurrentAccurals(credit);
assertEquals(expectedInterest, accruals.getAccount(INTEREST).get());
}
Testing complex past due scenarios:
@Test
public void pastDue1() {
// Setup credit with insufficient payment
charge(creditId, chargeDate, principal);
registerPayment(creditId, paymentDate, insufficientPayment);
// Calculate beyond payment due date
calculate(creditId, startDate, TODAY.plusMonths(1));
// Verify past due balances and late fee accruals
assertEquals(expectedPastDueInterest, credit.getActualSnapshot().getDebt().getAccount(PAST_DUE_INTEREST).get());
assertEquals(expectedLateFee, accrualService.calculateCurrentAccurals(credit).getAccount(LATE_FEE).get());
}
Testing Utilities
The test class provides reusable utility methods for common testing scenarios:
Credit Initialization
public UUID initCredit(LocalDate startDate, LocalDate today) {
return transactionTemplateBuilder.requiresNew().execute(s -> {
// Create complete credit structure: product, condition, application, credit
// Return credit ID for use in tests
});
}
Operation Creation
public void charge(UUID creditId, LocalDate operationDate, MonetaryAmount amount) {
transactionTemplateBuilder.requiresNew().executeWithoutResult(status -> {
chargeOperationService.createOperation(creditId, operationDate, amount);
});
}
public void registerPayment(UUID creditId, LocalDate paymentDate, MonetaryAmount amount) {
transactionTemplateBuilder.requiresNew().executeWithoutResult(status -> {
entityManager.find(ExampleCredit.class, creditId).getOperations()
.add(new ExampleCreditPayment(paymentDate, amount));
});
}
Test Data Management
The tests use realistic financial data and calculations:
private static final BigDecimal INTEREST_RATE = BigDecimal.valueOf(12); // 12% annual
private static final BigDecimal LATE_FEE_RATE = BigDecimal.valueOf(24); // 24% annual
private static final BigDecimal PRINCIPAL = BigDecimal.valueOf(2_000_000); // 2M ZWL
Interest calculations use proper day-count methods:
MonetaryAmount interest = principal.multiply(
(INTEREST_RATE.doubleValue() / 100d) * (ChronoUnit.DAYS.between(chargeDate, TODAY) / 360d)
);
Assertion Strategies
The tests demonstrate comprehensive assertion patterns:
Balance Verification
assertEquals(expectedAmount, credit.getActualSnapshot().getDebt().getAccount(PRINCIPAL).get());
Operation Effect Verification
ExampleCreditPayment payment = credit.getOperations(ExampleCreditPayment.class, APPROVED).findAny().get();
assertEquals(paymentAmount.negate(), payment.getFinalDebt().get().getAccount(INTEREST).get());
Best Practices for Operation Testing
-
Use Realistic Data: Test with actual financial amounts and rates that reflect real-world scenarios
-
Test Operation Sequences: Verify that operations work correctly when applied in different orders
-
Verify Time-based Calculations: Ensure accruals calculate correctly across different time periods
-
Test Edge Cases: Include scenarios like overpayments, zero balances, and boundary conditions
-
Use Proper Transactions: Each operation should be in its own transaction to mirror production behavior
-
Verify All Accounts: Check not just the primary effects but also secondary account impacts
-
Test Calculation Accuracy: Use precise mathematical calculations to verify financial accuracy
The testing approach in the example project provides a solid foundation for ensuring operation correctness and can be extended for custom operation types and business scenarios.
The Credit Operations Framework provides a solid foundation for building sophisticated financial applications while maintaining flexibility for customization and extension. By following these patterns and best practices, you can create robust, scalable operation processing systems that meet your specific business requirements.
24. Payment Transactions Framework
The Payment Transactions Framework connects real-world payments to credit operations. When someone makes a payment, it creates a transaction record, processes it through a payment gateway, and then creates the corresponding credit operation. This ensures every payment operation can be traced back to an actual payment.
24.1. The Basic Concept
Think of payment transactions as a receipt system:
-
Customer pays → Create transaction record (like writing a receipt)
-
Process payment → Send to bank/payment processor
-
Payment succeeds → Create credit operation (update the loan balance)
-
Keep records → Maintain complete audit trail
This two-step process (transaction → operation) ensures every balance change has a real-world payment behind it.
24.2. Architecture Overview
The system has three main layers that work together:
Transaction Layer
What it does: Records and processes real payments
BorrowerTransaction → PaymentGateway → Bank/Card Processor
Bridge Layer
What it does: Converts successful payments into credit operations
BorrowerTransactionService.handle() → Creates CreditOperation
Credit Layer
What it does: Updates loan balances and calculates interest
ExampleCreditPayment → Credit Calculation → Updated Balances
Core Components
The payment transaction system includes:
-
PaymentTransaction- Base record for all payment attempts -
BorrowerTransaction- Example project’s payment transaction type -
PaymentGateway- Interface for connecting to payment processors -
PaymentMethod- How customer wants to pay (bank account, card, etc.) -
BorrowerTransactionService- Converts successful payments to credit operations
24.3. Real-World Basis Principle
All operations in the credit system must have verifiable real-world foundations:
Payment Operations Foundation
Payment operations must originate from actual payment transactions:
Real Payment → PaymentTransaction → Success Handler → CreditOperation → Balance Update
Other Operations Foundations
While payment operations require transactions, other operations have their own real-world basis:
-
Accrual Operations → Contract terms and offer conditions (interest rates, payment schedules)
-
Past Due Operations → Payment schedule agreements (contractual due dates)
-
Charge Operations → Should be based on disbursements or outgoing transactions
|
The example project may have a flaw with charge operations - they should typically be tied to disbursement transactions, outgoing transactions, or regulatory events rather than being manually created without clear business justification. |
This principle ensures:
-
Regulatory Compliance - Every financial change can be traced to a real event
-
Audit Trail Integrity - Complete documentation of why changes occurred
-
Business Logic Accuracy - Operations reflect actual business events
-
Fraud Prevention - Prevents unauthorized or fictitious transactions
24.4. Transaction Entity Structure
Every payment attempt creates a PaymentTransaction record that tracks the payment from start to finish.
BorrowerTransaction Example
The example project uses BorrowerTransaction for loan payments:
@Entity
@DiscriminatorValue("BORROWER")
public class BorrowerTransaction extends PaymentTransaction {
@ManyToOne(fetch = FetchType.LAZY)
private ExampleCredit credit;
@ManyToOne(fetch = FetchType.LAZY)
@NotAudited
private CreditOperation operation;
public BorrowerTransaction(TransactionType type, MonetaryAmount amount,
PaymentMethod paymentMethod, ExampleCredit credit) {
super(type, amount);
this.credit = credit;
setPaymentMethod(paymentMethod);
}
public ExampleCredit getCredit() { return credit; }
public CreditOperation getOperation() { return operation; }
public void setOperation(CreditOperation operation) { this.operation = operation; }
}
Key fields:
-
credit- Which loan this payment is for -
operation- The credit operation created when payment succeeds -
amount- How much money (usesMonetaryAmountfor currency handling) -
type-INCOMING(customer pays) orOUTGOING(refund/disbursement) -
status- Current state of the payment -
paymentMethod- How the customer is paying (bank account, card, etc.)
Transaction Status Lifecycle
Transactions go through these states:
DRAFT → READY_FOR_EXECUTION → IN_PROGRESS → SUCCEED
↓ ↓ ↓ ↓
CANCELLED CANCELLED FAILED (Create Operation)
↓
UNAVAILABLE
↓
(Manual Review)
Understanding Transaction Status
The TransactionStatus enum has three important properties:
public enum TransactionStatus {
SUCCEED(800, true, true), // successful=true, complete=true
FAILED(700, false, true), // successful=false, complete=true
IN_PROGRESS(400, false, false); // successful=false, complete=false
private boolean successful; // Did the payment work?
private boolean complete; // Is processing finished?
}
Status meanings:
-
DRAFT- Transaction created, not yet sent to payment processor -
IN_PROGRESS- Sent to payment processor, waiting for response -
SUCCEED- Payment processor approved the payment -
FAILED- Payment processor declined the payment -
UNAVAILABLE- System error or payment processor is down -
CHARGEBACK- Bank reversed a previously successful payment
Transaction Processing Flow
The complete transaction processing follows this pattern:
-
Transaction Creation:
BorrowerTransactionentity created with payment details -
Gateway Submission:
PaymentTransactionServicesubmits to appropriate gateway -
Async Processing: Transaction processed asynchronously to avoid blocking
-
Status Updates: Transaction status updated based on gateway response
-
Success Handling:
BorrowerTransactionService.handle()creates credit operation -
Error Handling: Failed transactions trigger appropriate error responses
24.5. Payment Gateways
Payment gateways connect your system to banks and payment processors. Think of them as translators that convert your payment requests into the specific format each processor expects.
The Gateway Interface
All payment gateways implement the same interface:
public interface PaymentGateway {
String getMethodType(); // "ACH", "CARD", etc.
String getName(); // "Stripe", "Bank_ACH", etc.
boolean verify(PaymentMethod method) throws IOException;
TransactionResult proceedIncoming(String orderId, PaymentMethod method, MonetaryAmount amount);
TransactionResult proceedOutgoing(String orderId, PaymentMethod method, MonetaryAmount amount);
}
What each method does:
-
verify()- Validate payment method before processing (called byPaymentTransactionService.verify()) -
proceedIncoming()- Process customer payments (money coming in) -
proceedOutgoing()- Process refunds and disbursements (money going out)
Gateway Implementation Patterns
Payment gateways can be implemented following these patterns:
For immediate credit/debit card processing:
@Service
public class CardPaymentGateway implements PaymentGateway {
public TransactionResult proceedIncoming(String orderId, PaymentMethod method, MonetaryAmount amount) {
// 1. Extract payment method details (tokenized)
// 2. Build API request with transaction data
// 3. Submit to payment processor via HTTPS
// 4. Parse response and map to TransactionResult
// 5. Return standardized result with gateway reference
}
}
Key characteristics:
-
Immediate Processing - Real-time API calls with instant responses
-
Token-Based Security - Uses tokenized payment methods for PCI compliance
-
Structured Response - JSON/XML responses parsed into standard result format
-
Error Detection - Handles duplicate transactions and various error conditions
For traditional banking integration:
@Service
public class ACHGateway implements PaymentGateway {
public TransactionResult proceedOutgoing(String orderId, PaymentMethod method, MonetaryAmount amount) {
// 1. Build SOAP command with ACH details
// 2. Add merchant credentials and security headers
// 3. Submit via SOAP web service
// 4. Handle asynchronous ACH processing status
// 5. Return result with settlement timing information
}
}
Key characteristics:
-
SOAP Integration - XML-based web service communication
-
Asynchronous Processing - ACH transactions require settlement time
-
Comprehensive Logging - Full request/response logging for audit
-
Credential Management - Secure handling of merchant credentials
For bulk ACH processing via NACHA files:
@Service
public class NACHABatchGateway implements PaymentGateway {
@Scheduled(fixedRate = 3600000) // Hourly batch processing
public void processBatch() {
// 1. Find transactions ready for batch processing
// 2. Create NACHA-compliant file format
// 3. Add each transaction to appropriate batch
// 4. Generate file and transmit via SFTP
// 5. Update transaction statuses
}
}
Key characteristics:
-
Batch Processing - Multiple transactions in single file
-
File-Based Transport - SFTP or similar file delivery
-
NACHA Compliance - Proper ACH file format generation
-
Delayed Settlement - Transactions marked successful when file sent, not when settled
Gateway Configuration
Different gateways can be configured for different payment types:
@Service
public class ACHGateway implements PaymentGateway {
public String getMethodType() { return "ACH"; }
public String getName() { return "Bank_ACH"; }
}
@Service
public class CardGateway implements PaymentGateway {
public String getMethodType() { return "CARD"; }
public String getName() { return "Stripe"; }
}
The system selects the appropriate gateway based on the payment method type.
24.6. Payment Methods
Payment methods represent how customers want to pay - bank account, credit card, etc. Each payment method stores the necessary information to process payments through the appropriate gateway.
Example: LiquidityClientPaymentMethod
The example project includes a simple payment method for testing:
@Entity
@DiscriminatorValue(LiquidityClientPaymentMethod.TYPE)
public class LiquidityClientPaymentMethod extends PaymentMethod {
public static final String TYPE = LiquidityPaymentGateway.GATEWAY_TYPE;
@Column(name = "processed_date")
private LocalDate processedDate;
@Embedded
private MonetaryAmount amount;
@Column(name = "name")
private String ownerName;
public LiquidityClientPaymentMethod(LocalDate processedDate, MonetaryAmount amount,
TransactionType transactionType, String ownerName) {
super(TYPE);
this.processedDate = processedDate;
this.amount = amount;
this.transactionType = transactionType;
this.ownerName = ownerName;
}
public LocalDate getProcessedDate() { return processedDate; }
public MonetaryAmount getAmount() { return amount; }
public String getOwnerName() { return ownerName; }
}
This payment method:
-
Stores an amount - For testing, it has a fixed amount
-
Has a processed date - When the "payment" was processed
-
Works with gateways - Can be used by payment gateways that support this type
Payment Method Types
Different payment method types serve different use cases:
| Type | Use Case | Processing Pattern | Security Model |
|---|---|---|---|
ACH |
Bank account transfers |
Batch or real-time |
Account number encryption |
Debit/Credit Cards |
Card payments |
Real-time API |
PCI tokenization |
Digital Wallets |
Mobile payments |
Real-time API |
OAuth tokens |
Wire Transfers |
Large amounts |
Manual processing |
Bank verification |
Payment Method Implementation Patterns
When implementing new payment method types:
ACH Payment Method Pattern
@Entity
@DiscriminatorValue("ACH")
public class ACHPaymentMethod extends PaymentMethod {
// Encrypted bank account details
private String ownerName;
private String accountNumber; // Encrypted
private String routingNumber;
private AccountType accountType; // CHECKING, SAVINGS
// Validation and security methods
public boolean isValid() {
return validateRoutingNumber() && validateAccountNumber();
}
}
Key features:
-
Bank Account Details - Routing and account numbers for ACH processing
-
Account Type Classification - Checking vs savings account handling
-
Validation Logic - Routing number format and account number validation
-
Encryption - Sensitive account data encrypted at rest
Card Payment Method Pattern
@Entity
@DiscriminatorValue("CARD")
public class CardPaymentMethod extends PaymentMethod {
// Tokenized card data - no sensitive information stored
private String token; // From payment processor
private String lastFourDigits; // For display only
private String expiryMonth;
private String expiryYear;
public boolean isExpired() {
return LocalDate.now().isAfter(getExpiryDate());
}
}
Key characteristics:
-
Tokenization - Card numbers replaced with secure tokens from payment processor
-
PCI Compliance - No sensitive card data stored in application database
-
Display Information - Only last four digits stored for user interface
-
Expiry Validation - Built-in expiration checking
Payment Method Security
The framework implements comprehensive security patterns:
Data Protection
// Sensitive data encrypted at rest
@Convert(converter = EncryptedStringConverter.class)
private String accountNumber;
// Tokens from external processors
private String processorToken;
// Display-only information
private String maskedAccountNumber; // "****1234"
Validation and Verification
public interface PaymentMethodValidator {
boolean validate(PaymentMethod method);
ValidationResult verify(PaymentMethod method) throws IOException;
}
// Gateway-specific validation
@Override
public boolean verify(PaymentMethod method) throws IOException {
// Real-time validation with payment processor
return gateway.validatePaymentMethod(method);
}
24.7. How Transactions Become Operations
When a payment succeeds, the system needs to update the loan balance. This happens in BorrowerTransactionService.handle().
The Conversion Process
Here’s what happens when a payment succeeds:
@Override
public void handle(PaymentTransaction t) {
BorrowerTransaction transaction = (BorrowerTransaction) t;
if (transaction.getStatus() == TransactionStatus.SUCCEED) {
ExampleCredit credit = transaction.getCredit();
LocalDate date = getProcessedDate(transaction);
// Create the right type of operation
CreditOperation operation = switch (transaction.getType()) {
case INCOMING -> handleIncoming(credit, transaction, date); // Customer payment
case OUTGOING -> handleOutgoing(credit, transaction, date); // Refund/disbursement
};
// Link them together for audit trail
transaction.setOperation(operation);
}
}
Customer Payments (INCOMING)
When a customer makes a payment:
private CreditPayment handleIncoming(ExampleCredit credit, BorrowerTransaction transaction, LocalDate date) {
// Create payment operation
CreditPayment payment = new ExampleCreditPayment(date, transaction.getAmount());
// Register with credit system
return paymentService.registerPayment(credit, payment);
}
This creates an ExampleCreditPayment operation that reduces the loan balance.
Refunds and Disbursements (OUTGOING)
When money goes to the customer:
private ChargeOperation handleOutgoing(ExampleCredit credit, BorrowerTransaction transaction, LocalDate date) {
// Create charge operation (increases balance)
return chargeOperationService.createOperation(credit.getId(), date, transaction.getAmount());
}
This creates a ChargeOperation that increases the loan balance (for disbursements) or reverses payments (for refunds).
The Audit Trail
The system maintains complete traceability:
-
Transaction Record - Shows the real-world payment attempt
-
Gateway Response - Stored in
transaction.tracefield -
Operation Link -
transaction.operationpoints to the credit operation -
Credit Update - Operation appears in credit’s operation list
This means you can always trace a balance change back to the original payment.
24.8. Processing Payments Asynchronously
Payment processing happens in the background so users don’t have to wait. When someone submits a payment, the system:
-
Creates transaction record - Saves it immediately
-
Returns to user - Shows "processing" message
-
Processes in background - Calls payment gateway
-
Updates status - Success or failure
-
Creates operation - If payment succeeded
Why Async Processing?
-
Faster user experience - Don’t wait for slow payment processors
-
Better error handling - Can retry failed payments
-
Scalability - Handle many payments at once
Error Handling
When processing payments, three things can happen:
Payment Declined
// Gateway says "insufficient funds" or "invalid card"
transaction.setStatus(TransactionStatus.FAILED);
transaction.addTrace("Gateway declined: " + result.getMessage());
System Error
// Code bug or unexpected error
transaction.setStatus(TransactionStatus.UNAVAILABLE);
transaction.addTrace("System error: " + e.getMessage());
Gateway Down
// Payment processor is unavailable
transaction.setStatus(TransactionStatus.UNAVAILABLE);
transaction.addTrace("Gateway unavailable: " + e.getMessage());
// Can retry later
24.9. Transaction Types and Patterns
Different transaction types serve different business purposes and follow specific processing patterns.
Incoming Payment Transactions
Borrower payments to reduce credit balances:
BorrowerTransaction payment = new BorrowerTransaction(
TransactionType.INCOMING,
credit,
paymentMethod,
paymentAmount,
"Borrower payment"
);
Processing flow:
-
User Initiates - Borrower submits payment through portal
-
Transaction Created -
BorrowerTransactionentity persisted -
Gateway Processing - Payment method charged via appropriate gateway
-
Success Handling -
ExampleCreditPaymentoperation created -
Balance Update - Credit calculation applies payment to debt accounts
Outgoing Payment Transactions
Disbursements or refunds to borrowers:
BorrowerTransaction disbursement = new BorrowerTransaction(
TransactionType.OUTGOING,
credit,
paymentMethod,
disbursementAmount,
"Loan disbursement"
);
Processing flow:
-
System Initiates - Loan approval triggers disbursement
-
Transaction Created - Outgoing transaction entity
-
Gateway Processing - Funds sent to borrower account
-
Success Handling - Disbursement operation created
-
Balance Update - Principal balance increased
Chargeback Transactions
Handling payment reversals:
// Original payment is reversed
originalTransaction.setStatus(TransactionStatus.CHARGEBACK);
// Chargeback operation created to reverse the payment
ChargebackOperation chargeback = new ChargebackOperation(
originalPayment.getAmount().negate(),
"Chargeback: " + originalTransaction.getOrderId()
);
Retry Patterns
Failed transactions may be retried based on failure type:
if (canRetry(transaction, result)) {
scheduleRetry(transaction, calculateBackoffDelay(transaction.getRetryCount()));
} else {
markPermanentFailure(transaction, result);
}
Retry logic considers:
-
Failure Type - Network errors retryable, declines usually not
-
Retry Count - Exponential backoff with maximum attempts
-
Time Limits - Don’t retry indefinitely old transactions
24.10. Testing Payment Transactions
Testing payment transactions requires careful consideration of external dependencies and asynchronous processing.
Test Gateway Implementation
For testing, implement a controllable test gateway:
@Service
public class TestPaymentGateway implements PaymentGateway {
@Override
public TransactionResult proceedIncoming(String orderId, PaymentMethod method, MonetaryAmount amount) {
// Simulate different scenarios based on test data
if (amount.getNumber().doubleValue() == 999.99) {
return new TransactionResult(orderId, amount, Status.FAIL, false, "Test decline");
}
if ("ERROR_TOKEN".equals(method.getToken())) {
throw new RuntimeException("Test gateway error");
}
return new TransactionResult(orderId, amount, Status.SUCCESS, false, "Test success");
}
}
Integration Testing Patterns
Test the complete transaction-to-operation flow:
@Test
@Transactional
public void testSuccessfulPayment() {
// Setup: Create credit and payment method
UUID creditId = createTestCredit();
PaymentMethod paymentMethod = createTestPaymentMethod();
// Execute: Process payment transaction
BorrowerTransaction transaction = new BorrowerTransaction(
TransactionType.INCOMING, credit, paymentMethod,
MonetaryAmount.of(500, "USD"), "Test payment"
);
paymentTransactionService.processTransaction(transaction.getId());
// Wait for async processing
await().atMost(5, SECONDS).until(() ->
transactionRepository.findById(transaction.getId()).getStatus() == TransactionStatus.SUCCEED
);
// Verify: Check operation was created and credit updated
ExampleCredit updatedCredit = creditRepository.findById(creditId);
assertThat(updatedCredit.getOperations(ExampleCreditPayment.class)).hasSize(1);
ExampleCreditPayment payment = updatedCredit.getOperations(ExampleCreditPayment.class).iterator().next();
assertThat(payment.getAmount()).isEqualTo(MonetaryAmount.of(500, "USD"));
assertThat(payment.getTransaction()).isEqualTo(transaction);
}
Mocking External Dependencies
For unit tests, mock gateway dependencies:
@MockBean
PaymentGateway mockGateway;
@Test
public void testGatewayFailure() {
// Setup: Mock gateway to return failure
when(mockGateway.proceedDebit(any(), any(), any()))
.thenReturn(new TransactionResult("123", amount, Status.FAIL, false, "Declined"));
// Execute: Process transaction
paymentTransactionService.processTransaction(transactionId);
// Verify: Transaction marked as failed
BorrowerTransaction transaction = transactionRepository.findById(transactionId);
assertThat(transaction.getStatus()).isEqualTo(TransactionStatus.FAILED);
// Verify: No operation created
assertThat(credit.getOperations()).isEmpty();
}
24.11. Security and Compliance
Payment transaction processing requires adherence to strict security and compliance standards.
PCI DSS Compliance
For card payments:
-
No Card Storage - Card numbers never stored in application database
-
Tokenization - Sensitive data replaced with non-sensitive tokens
-
Secure Transmission - All payment data encrypted in transit
-
Access Controls - Role-based access to payment functionality
Bank Security Standards
For ACH payments:
-
Encryption at Rest - Bank account data encrypted in database
-
Secure APIs - TLS encryption for all gateway communication
-
Credential Management - Secure storage of gateway credentials
-
Audit Logging - Complete transaction audit trails
Regulatory Compliance
Financial regulations require:
-
Transaction Traceability - Complete audit trail from user action to balance change
-
Data Retention - Transaction records maintained for required periods
-
Reporting - Transaction data available for regulatory reporting
-
Error Handling - Proper handling and reporting of failed transactions
24.12. Best Practices
Transaction Design Principles
-
Idempotency - Transactions should produce same result when retried
-
Atomicity - Each transaction represents a single business event
-
Traceability - Complete audit trail from initiation to completion
-
Error Recovery - Graceful handling of all failure scenarios
Gateway Integration Best Practices
-
Timeout Handling - Appropriate timeouts for gateway calls
-
Retry Logic - Intelligent retry strategies for transient failures
-
Rate Limiting - Respect gateway rate limits and quotas
-
Monitoring - Comprehensive monitoring of gateway performance
24.13. Summary
The Payment Transactions Framework ensures every credit operation has a real-world basis:
The Flow:
Customer Payment → BorrowerTransaction → PaymentGateway → Bank/Processor
↓
Payment Succeeds → BorrowerTransactionService.handle() → ExampleCreditPayment → Updated Loan Balance
Key Benefits:
-
Complete audit trail - Every balance change traces to a real payment
-
Async processing - Fast user experience, reliable background processing
-
Multiple gateways - Support different payment processors
-
Error handling - Graceful handling of declined payments and system errors
-
Regulatory compliance - Full documentation for audits
For Developers:
-
Extend
PaymentTransactionfor your transaction types -
Implement
PaymentGatewayfor new payment processors -
Use
PaymentTransactionHandlerto convert transactions to operations -
Always maintain the transaction → operation link for audit trails
This foundation supports any type of payment processing while ensuring complete traceability and regulatory compliance.