Loan Origination System (LOS)

A robust backend system that safely and concurrently processes multiple loan applications with automated system approval delays, agent-manager hierarchies, and mock notifications. Built with Spring Boot 3 and MySQL.

🎯 Overview

This system demonstrates enterprise-level backend development skills by implementing:

• Multithreaded processing with 25-second delay simulation
• Automated decision-making based on customizable business rules
• Agent-manager hierarchy with automatic assignment workflow
• Mock notification system for push and SMS alerts
• Thread-safe concurrent processing using ReentrantLock
• Real-time monitoring and analytics

🛠️ Technology Stack

• Backend: Spring Boot 3.5.5
• Language: Java 17
• Database: MySQL 8.0
• ORM: Hibernate/JPA
• Build Tool: Maven
• Architecture: Layered (Controller → Service → Repository → Entity)

📋 Prerequisites

• Java 17 or higher
• MySQL 8.0 or higher
• Maven 3.6 or higher
• IDE (IntelliJ IDEA, Eclipse, or VS Code)

🚀 Quick Setup

1. Database Configuration

CREATE DATABASE loan_origination_db;

2. Application Configuration

Update src/main/resources/application.yml:

spring:
  datasource:
    url: jdbc:mysql://localhost:3306/loan_origination_db
    username: root
    password: password

3. Insert Sample Data

INSERT INTO agents (agent_name, email, phone, is_available, manager_id, created_at) VALUES
('Rajesh Kumar Sharma', 'rajesh.sharma@bank.com', '9876543210', true, null, NOW()),
('Priya Gupta', 'priya.gupta@bank.com', '9876543211', true, 1, NOW()),
('Amit Patel', 'amit.patel@bank.com', '9876543212', true, 1, NOW()),
('Sunita Verma', 'sunita.verma@bank.com', '9876543213', true, 2, NOW());

4. Run Application

mvn spring-boot:run

Application URL: http://localhost:8080

📖 API Endpoints

Endpoint	Method	Description
/api/v1/loans	POST	Submit new loan application
/api/v1/loans/status-count	GET	Get real-time status counts
/api/v1/loans	GET	Get loans by status (paginated)
/api/v1/customers/top	GET	Get top 3 customers with most approved loans
/api/v1/agents/{agentId}/loans/{loanId}/decision	PUT	Agent makes approve/reject decision

🧪 Sample Usage

Submit Loan Application

POST http://localhost:8080/api/v1/loans
Content-Type: application/json

{
  "loanId": "LOAN001",
  "customerName": "Ramesh Kumar Singh",
  "customerPhone": "9123456789",
  "loanAmount": 250000,
  "loanType": "PERSONAL"
}

Check Status Counts

GET http://localhost:8080/api/v1/loans/status-count

Response:

{
  "applied": 0,
  "approvedBySystem": 9,
  "rejectedBySystem": 16,
  "underReview": 16,
  "approvedByAgent": 0,
  "rejectedByAgent": 0
}

Agent Decision

PUT http://localhost:8080/api/v1/agents/1/loans/LOAN001/decision
Content-Type: application/json

{
  "decision": "APPROVE"
}

🔄 Loan Processing Workflow

1. Application Submission → Status: APPLIED

2. Background Processing (25-second delay) → Business rules applied:

   • Auto-approve: Personal loans ≤ ₹50,000 → APPROVED_BY_SYSTEM
   • Auto-reject: Any loan > ₹50,00,000 → REJECTED_BY_SYSTEM
   • Manual review: Other cases → UNDER_REVIEW

3. Agent Assignment → Available agent assigned automatically

4. Notifications → Push to agent, SMS to customer

5. Agent Decision → APPROVED_BY_AGENT or REJECTED_BY_AGENT

📊 Business Rules

Condition	Action	Result Status
Personal loan ≤ ₹50,000	Auto-approve	APPROVED_BY_SYSTEM
Any loan > ₹50,00,000	Auto-reject	REJECTED_BY_SYSTEM
Other amounts	Random decision	30% approve, 30% reject, 40% review

🧪 Testing

Using Postman Collection

Import LOS-API-Collection.json for comprehensive testing of all endpoints.

Test Data Samples

// Auto-approve scenario
{
  "loanId": "AUTO_APPROVE_001",
  "customerName": "Test User 1",
  "customerPhone": "9999999991",
  "loanAmount": 45000,
  "loanType": "PERSONAL"
}

// Manual review scenario
{
  "loanId": "MANUAL_REVIEW_001",
  "customerName": "Test User 2",
  "customerPhone": "9999999992",
  "loanAmount": 250000,
  "loanType": "BUSINESS"
}

// Auto-reject scenario
{
  "loanId": "AUTO_REJECT_001",
  "customerName": "Test User 3",
  "customerPhone": "9999999993",
  "loanAmount": 6000000,
  "loanType": "BUSINESS"
}

🏗️ Architecture

src/
├── main/
│   ├── java/com/example/LOS/
│   │   ├── entity/
│   │   │   ├── LoanApplication.java
│   │   │   └── Agent.java
│   │   ├── repository/
│   │   │   ├── LoanApplicationRepository.java
│   │   │   └── AgentRepository.java
│   │   ├── service/
│   │   │   ├── LoanApplicationService.java
│   │   │   ├── LoanProcessingService.java
│   │   │   ├── ScheduledLoanProcessor.java
│   │   │   ├── impl/
│   │   │   │   └── LoanApplicationServiceImpl.java
│   │   │   └── notification/
│   │   │       ├── NotificationService.java
│   │   │       └── MockNotificationService.java
│   │   ├── controller/
│   │   │   ├── LoanApplicationController.java
│   │   │   └── AgentController.java
│   │   ├── dto/
│   │   │   ├── LoanApplicationRequest.java
│   │   │   ├── LoanApplicationResponse.java
│   │   │   ├── AgentDecisionRequest.java
│   │   │   ├── StatusCountResponse.java
│   │   │   └── TopCustomerResponse.java
│   │   ├── enums/
│   │   │   ├── ApplicationStatus.java
│   │   │   └── LoanType.java
│   │   ├── config/
│   │   │   └── ThreadPoolConfig.java
│   │   ├── exception/
│   │   │   ├── GlobalExceptionHandler.java
│   │   │   ├── ResourceNotFoundException.java
│   │   │   └── InvalidStatusTransitionException.java
│   │   └── LoanOriginationSystemApplication.java
│   └── resources/
│       ├── application.yml
│       └── data.sql
└── test/
    └── java/com/example/LOS/
        ├── service/
        │   └── LoanApplicationServiceTest.java
        └── controller/
            └── LoanApplicationControllerTest.java

✅ Features Implemented

• ✅ Multithreaded Processing: Background jobs with thread pool
• ✅ Thread Safety: ReentrantLock prevents race conditions
• ✅ Automated Decision Making: Business rule engine
• ✅ Agent Workflow: Hierarchical assignment and notifications
• ✅ Mock Services: Push and SMS notification simulation
• ✅ Real-time Analytics: Status monitoring and top customers
• ✅ Pagination Support: Efficient data retrieval
• ✅ Input Validation: Comprehensive request validation
• ✅ Error Handling: Global exception management
• ✅ Clean Architecture: Layered design with separation of concerns

🔧 Configuration

Thread Pool Settings

• Core threads: 3
• Max threads: 10
• Queue capacity: 50
• Processing interval: Every 5 seconds
• Simulation delay: 25 seconds per loan

Database Optimization

• Indexed columns: application_status, customer_name
• Connection pooling: HikariCP
• Transaction management: Spring @Transactional

📈 Performance Features

• Concurrent Processing: Multiple loans processed simultaneously
• Database Indexing: Optimized query performance
• Async Operations: Non-blocking background processing
• Memory Efficient: Pagination for large datasets
• Connection Pooling: Efficient database resource usage

🎯 Assignment Requirements Fulfilled

Requirement	Status	Implementation
Submit Loan Application	✅	POST /api/v1/loans with validation
Multithreaded Processing	✅	Background jobs with 25-second delay
Agent Assignment & Review	✅	Automatic assignment with hierarchy
Mock Notifications	✅	Push and SMS services
Status Monitoring	✅	Real-time counts endpoint
Top Customers API	✅	Analytics with approved loan counts
Pagination Support	✅	GET /api/v1/loans with paging
Thread Safety	✅	ReentrantLock implementation
Clean Architecture	✅	Layered design pattern
Database Design	✅	Normalized schema with indexes

---

Built with Spring Boot 3.5.5 & Java 17

This project demonstrates enterprise-grade loan processing capabilities with concurrent handling, automated workflows, and comprehensive monitoring suitable for production banking systems.