Credit Risk Analysis Batch

This Spring Batch application is designed to analyze and evaluate customer credit risk using a sophisticated multifactorial scoring system. The system processes customer data from a CSV file, performs detailed risk analysis, and generates results in both an H2 database and an output CSV file.

• Functional tests video

Index 📜

View details

Latest update: 19/02/26

Section 1) Description, configuration and technologies

• 1.0) Project Description.
• 1.1) Main Features.
• 1.2) Configuration and Execution.
• 1.3) Technologies.

Section 2) Data Format, Use Cases and Results

• 2.0) Input Data Format.
• 2.1) Input File Example.
• 2.2) Use Cases.
• 2.3) Analysis Factors.
• 2.4) Results and Monitoring.

Section 3) Functionality Testing and References

• 3.0) Functionality Test.
• 3.1) Contribution and License.

Section 1) Description, configuration and technologies

1.0) Project Description 🔝

View details

This Spring Batch application is designed to analyze and evaluate customer credit risk using a sophisticated multifactorial scoring system. The system processes customer data from a CSV file, performs detailed risk analysis, and generates results in both an H2 database and an output CSV file. It uses a base credit score normalization, multiple risk factors (income, debt-to-income ratio, payment history), a sigmoid function for default probability calculation, and risk categorization into LOW, MEDIUM, HIGH, and VERY_HIGH levels.

The application also includes a recommendation system that provides suggested credit limits, recommended interest rates, credit terms, and specific recommendations by risk factor. Batch processing is handled through configurable chunk processing with error handling and retries. Monitoring is enabled via Actuator endpoints, Prometheus integration, detailed logging, and performance metrics.

1.1) Main Features 🔝

View details

Multifactorial Risk Analysis

• Base Credit Score: Credit score normalization and weighting
• Risk Factors:
  • Income Factor
  • Debt-to-Income Ratio Factor
  • Payment History Factor
• Default Probability Calculation: Using sigmoid function
• Risk Categorization: LOW, MEDIUM, HIGH, VERY_HIGH

Recommendation System

• Suggested credit limits
• Recommended interest rates
• Credit terms
• Specific recommendations by risk factor

Batch Processing

• CSV data reading
• Configurable chunk processing
• Database and CSV file writing
• Error handling and retries

Monitoring and Metrics

• Actuator endpoints for monitoring
• Prometheus integration
• Detailed logging
• Performance metrics

1.2) Configuration and Execution 🔝

View details

• We create a work environment through some IDE, we may or may not create a root folder for the project, we position ourselves on it

cd 'projectRootName'

• Once a work environment has been created, we clone the project

git clone https://github.com/andresWeitzel/Credit_Risk_Analysis_Batch

• We position ourselves on the project

cd 'Credit_Risk_Analysis_Batch'

• We install Java 17 or higher and Maven 3.6 or higher if not already installed.

• We build the project

mvn clean install

• We run the application

mvn spring-boot:run

• Important: It is possible that there are other previous steps that have not been included due to synchronization between docs in relation to development. Please open a conversation thread within the 'Issues' section of the project.

1.3) Technologies 🔝

View details

Technologies	Version	Purpose
Java	17+	Programming Language
Spring Boot	3.x	Application Framework
Spring Batch	5.x	Batch Processing Framework
H2 Database	2.x	In-memory Database
Caffeine Cache	3.x	High Performance Caching
Maven	3.6+	Build and Dependency Management
JUnit	5.x	Unit Testing Framework
Postman	10+	Http Client
Git	2.x	Version Control
Others	-	Others

Project Structure

src/main/java/com/example/creditrisk/
├── config/
│   └── BatchConfig.java         # Batch job configuration
├── model/
│   └── CreditRiskData.java      # Data model
├── processor/
│   └── CreditRiskProcessor.java # Risk processor
├── reader/
│   └── CreditRiskItemReader.java # Data reader
├── service/
│   └── RiskAnalysisService.java # Risk analysis service
├── writer/
│   └── CreditRiskFileWriter.java # Results writer
└── CreditRiskBatchApplication.java

Section 2) Data Format, Use Cases and Results

2.0) Input Data Format 🔝

View details

The input file must be in CSV format with the following columns:

Basic Customer Information

• customerId: Unique customer identifier
• customerName: Customer name
• birthDate: Birth date (YYYY-MM-DD)
• age: Customer age
• maritalStatus: Marital status (SINGLE, MARRIED, DIVORCED)
• educationLevel: Education level (HIGH_SCHOOL, BACHELORS, MASTERS, DOCTORATE)

Financial Information

• creditScore: Credit score (0-850)
• income: Annual income
• debtToIncomeRatio: Debt-to-income ratio (%)
• monthlyExpenses: Monthly expenses
• savingsBalance: Savings balance
• propertyValue: Property value (if applicable)

Employment Information

• employmentType: Employment type (FULL_TIME, PART_TIME)
• employmentYears: Years of employment
• industry: Industry/Sector

Credit History

• paymentHistory: Payment history (0-100)
• creditHistoryYears: Years of credit history
• numberOfCreditCards: Number of credit cards
• creditCardUtilization: Credit card utilization (%)
• hasBankruptcy: Bankruptcy indicator
• bankruptcyYearsAgo: Years since bankruptcy (if applicable)
• hasForeclosure: Foreclosure indicator
• foreclosureYearsAgo: Years since foreclosure (if applicable)

Loan Information

• loanAmount: Loan amount
• loanPurpose: Loan purpose (MORTGAGE, CAR, PERSONAL, BUSINESS, EDUCATION, CONSOLIDATION, HOME_IMPROVEMENT, INVESTMENT)
• existingLoans: Number of existing loans
• loanTerm: Loan term (SHORT_TERM, MEDIUM_TERM, LONG_TERM)
• interestRate: Interest rate (%)
• collateralType: Collateral type (NONE, REAL_ESTATE, VEHICLE, BUSINESS)
• collateralValue: Collateral value

Residence Information

• residenceType: Residence type (OWN, RENT)
• yearsAtCurrentAddress: Years at current address

Guarantor Information

• guarantorStatus: Guarantor status (NONE, REQUIRED)
• guarantorCreditScore: Guarantor credit score
• guarantorIncome: Guarantor income
• guarantorRelationship: Relationship with guarantor (FAMILY, FRIEND, NA)

Analysis Fields

• riskCategory: Risk category (LOW, MEDIUM, HIGH, VERY_HIGH)
• status: Application status (ACTIVE, INACTIVE)
• additionalInfo: Additional information
• recommendations: Specific recommendations

2.1) Input File Example 🔝

View details

customerId,customerName,creditScore,income,debtToIncomeRatio,paymentHistory,employmentYears,loanAmount,loanPurpose,existingLoans,propertyValue,maritalStatus,educationLevel,industry,riskCategory,status,additionalInfo,recommendations,birthDate,age,employmentType,monthlyExpenses,savingsBalance,creditHistoryYears,numberOfCreditCards,creditCardUtilization,hasBankruptcy,bankruptcyYearsAgo,hasForeclosure,foreclosureYearsAgo,residenceType,yearsAtCurrentAddress,loanTerm,interestRate,collateralType,collateralValue,guarantorStatus,guarantorCreditScore,guarantorIncome,guarantorRelationship
CUST001,John Smith,720,85000,0.35,95,8,150000,Mortgage,1,350000,Married,Bachelors,Technology,Low,Approved,Stable employment history,Consider refinancing in 2 years,1980-05-15,43,Full-time,3500,50000,15,2,0.25,false,0,false,0,Mortgage,5,Long-term,4.5,Real Estate,350000,None,,,,
CUST002,Maria Garcia,680,65000,0.45,88,5,120000,Car,2,250000,Single,Masters,Healthcare,Medium,Approved,Recent job change,Monitor debt ratio,1988-08-22,35,Full-time,2800,25000,8,3,0.45,false,0,false,0,Rent,2,Medium-term,5.2,Vehicle,45000,None,,,,
CUST003,Robert Johnson,580,45000,0.65,75,3,80000,Personal,3,180000,Divorced,High School,Retail,High,Rejected,Multiple late payments,Improve credit score,1975-11-30,48,Part-time,2200,5000,5,4,0.85,true,3,false,0,Rent,1,Short-term,7.5,None,0,Required,650,55000,Parent
CUST004,Sarah Williams,750,120000,0.25,98,12,200000,Business,0,500000,Married,PhD,Finance,Low,Approved,Excellent credit history,Consider investment opportunities,1972-03-10,51,Self-employed,4500,150000,20,2,0.15,false,0,false,0,Own,10,Long-term,4.0,Real Estate,500000,None,,,,
CUST005,Michael Brown,620,55000,0.55,82,4,95000,Education,1,220000,Single,Bachelors,Education,Medium,Approved,Student loan debt,Consolidate loans,1990-07-18,33,Full-time,2300,15000,6,2,0.60,false,0,false,0,Rent,3,Medium-term,5.8,None,0,None,,,,
CUST006,Lisa Chen,710,95000,0.30,92,7,180000,Home Improvement,2,400000,Married,Masters,Engineering,Low,Approved,High property value,Consider home equity,1985-12-05,38,Full-time,3800,75000,12,3,0.30,false,0,false,0,Mortgage,6,Long-term,4.2,Real Estate,400000,None,,,,
CUST007,David Miller,590,48000,0.60,78,2,70000,Debt Consolidation,4,150000,Single,Associates,Manufacturing,High,Rejected,High debt load,Debt management plan,1982-09-25,41,Contract,2100,8000,4,5,0.90,false,0,true,2,Rent,1,Short-term,8.0,None,0,Required,680,60000,Sibling
CUST008,Emma Wilson,730,110000,0.28,96,10,250000,Investment,1,600000,Married,Bachelors,Real Estate,Low,Approved,Investment property,Portfolio diversification,1978-04-12,45,Self-employed,4200,200000,18,2,0.20,false,0,false,0,Own,8,Long-term,4.8,Real Estate,600000,None,,,,
CUST009,James Taylor,650,72000,0.40,85,6,130000,Medical,2,280000,Married,Bachelors,Healthcare,Medium,Approved,Medical expenses,Health insurance review,1987-06-30,36,Full-time,3000,35000,9,3,0.50,false,0,false,0,Mortgage,4,Medium-term,5.5,Real Estate,280000,None,,,,
CUST010,Sophia Martinez,670,68000,0.42,87,5,115000,Wedding,1,260000,Engaged,Bachelors,Marketing,Medium,Approved,Upcoming wedding,Budget planning,1992-02-14,31,Full-time,2900,20000,7,2,0.40,false,0,false,0,Rent,2,Short-term,6.0,None,0,Provided,700,75000,Fiance

2.2) Use Cases 🔝

View details

Low Risk Customer (CUST001)

{
  "customerId": "CUST001",
  "riskCategory": "LOW",
  "baseScore": 88.2,
  "incomeRiskFactor": 0.2,
  "debtRiskFactor": 0.15,
  "paymentRiskFactor": 0.05,
  "finalScore": 85.0,
  "defaultProbability": 0.12,
  "recommendations": {
    "creditLimit": "Increase credit limit",
    "interestRate": "Offer preferred rate",
    "terms": "Flexible terms"
  }
}

Medium Risk Customer (CUST002)

{
  "customerId": "CUST002",
  "riskCategory": "MEDIUM",
  "baseScore": 80.0,
  "incomeRiskFactor": 0.35,
  "debtRiskFactor": 0.25,
  "paymentRiskFactor": 0.15,
  "finalScore": 65.0,
  "defaultProbability": 0.28,
  "recommendations": {
    "creditLimit": "Maintain current limit",
    "interestRate": "Standard rate",
    "terms": "Standard terms",
    "payment": "Suggest structured payment plan"
  }
}

High Risk Customer (CUST003)

{
  "customerId": "CUST003",
  "riskCategory": "HIGH",
  "baseScore": 68.2,
  "incomeRiskFactor": 0.55,
  "debtRiskFactor": 0.45,
  "paymentRiskFactor": 0.30,
  "finalScore": 45.0,
  "defaultProbability": 0.62,
  "recommendations": {
    "creditLimit": "Reduce credit limit",
    "interestRate": "Higher rate",
    "terms": "Stricter terms",
    "income": "Request additional income verification",
    "debt": "Recommend debt reduction",
    "payment": "Suggest structured payment plan"
  }
}

2.3) Analysis Factors 🔝

View details

Base Credit Score

• Normalization: (creditScore / 850) * 100
• Weight: 40% of final score
• Example: 750 points → 88.2 normalized points

Income Factor

• Normalization: min(income / 200000, 1.0)
• Weight: 30% of final score
• Example: $120,000 → 0.6 normalized

Debt-to-Income Ratio Factor

• Normalization: min(debtToIncomeRatio / 100, 1.0)
• Weight: 20% of final score
• Example: 0.35 (35%) → 0.35 normalized

Payment History Factor

• Normalization: paymentHistory / 100
• Weight: 10% of final score
• Example: 95% → 0.95 normalized

2.4) Results and Monitoring 🔝

View details

The application generates two types of output:

H2 Database

• Table: credit_risk_data
• Location: Memory (H2 in-memory database)
• Access: http://localhost:8080/h2-console
• Credentials:
  • JDBC URL: jdbc:h2:mem:creditriskdb
  • Username: sa
  • Password: (empty)

Output CSV File

• Location: src/main/resources/output/credit-risk-results.csv
• Format: CSV with the following columns:

Original Fields

• All fields from the input file

Analysis Fields

• baseScore: Normalized base score (0-100)
• incomeRiskFactor: Income risk factor (0-1)
• debtRiskFactor: Debt risk factor (0-1)
• paymentRiskFactor: Payment history risk factor (0-1)
• finalScore: Weighted final score (0-100)
• defaultProbability: Default probability (0-1)
• riskCategory: Risk category (LOW, MEDIUM, HIGH, VERY_HIGH)

Recommendation Fields

• creditLimitRecommendation: Credit limit recommendation
• interestRateRecommendation: Interest rate recommendation
• termsRecommendation: Terms recommendation
• incomeRecommendation: Income-related recommendation
• debtRecommendation: Debt-related recommendation
• paymentRecommendation: Payment-related recommendation

Monitoring Endpoints

The application exposes several monitoring endpoints:

• /actuator/health: Application status
• /actuator/metrics: Performance metrics
• /actuator/prometheus: Prometheus format metrics

Section 3) Functionality Testing and References.

3.0) Functionality Test 🔝

View details

Watch video

3.1) Contribution and License 🔝

View details

Contribution

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

License

This project is under the MIT License. See the LICENSE file for more details.