Skip to content

QUICKSTART.md

Latest commit

 

History

History
176 lines (136 loc) · 3.95 KB

QUICKSTART.md

File metadata and controls

176 lines (136 loc) · 3.95 KB

Quick Start Guide

Your Application is Already Running! 🚀

Both the backend and frontend are currently running:

How to Use the Application

Step 1: Open the Web Interface

Open your browser and navigate to:

http://localhost:8082

Step 2: Fill in the Credit Assessment Form

The form includes the following fields:

Applicant Profile:

  • Age (18-100)
  • Number of Dependents (0-10)

Financial Stability:

  • Annual Income ($)
  • Debt-to-Income Ratio (0-1, e.g., 0.45 for 45%)

Credit History:

  • Revolving Utilization (0-1, e.g., 0.6 for 60%)
  • Open Credit Lines (number of open accounts)
  • Delinquencies in Last 2 Years
  • FICO Score (300-850)

Loan Details:

  • Loan Amount ($)
  • Employment Length (years)

Step 3: Submit and View Results

  1. Click "Analyze Credit Risk"
  2. Wait for the prediction (should take 1-2 seconds)
  3. View the results:
    • Default Probability: The likelihood of loan default (0-100%)
    • Risk Label: LOW, MEDIUM, or HIGH
    • Top Risk Factors: The features that most influence the prediction

Example Test Data

Try these sample applicants:

Low Risk Applicant

Age: 45
Annual Income: $85,000
Debt-to-Income Ratio: 0.25
Revolving Utilization: 0.30
Open Credit Lines: 8
Delinquencies (2 yrs): 0
Dependents: 2
FICO Score: 780
Loan Amount: $15,000
Employment Length: 10 years

High Risk Applicant

Age: 28
Annual Income: $35,000
Debt-to-Income Ratio: 0.65
Revolving Utilization: 0.95
Open Credit Lines: 3
Delinquencies (2 yrs): 4
Dependents: 0
FICO Score: 580
Loan Amount: $30,000
Employment Length: 1 year

Testing via API (Advanced)

If you prefer to test via command line:

curl -X POST http://localhost:8000/api/predict \
  -H "Content-Type: application/json" \
  -d '{
    "age": 35,
    "annual_income": 60000,
    "debt_to_income_ratio": 0.45,
    "revolving_utilization": 0.6,
    "open_credit_lines": 5,
    "delinquencies_2yrs": 2,
    "dependents": 1,
    "fico_score": 720,
    "loan_amount": 25000,
    "employment_length": 5
  }'

Stopping the Application

Stop the Frontend

# Find the npm process
ps aux | grep "npm run dev"

# Kill it using the PID
kill <PID>

Stop the Backend

# Find the uvicorn process
ps aux | grep uvicorn

# Kill it using the PID
kill <PID>

Or simply press Ctrl+C in the terminal where each service is running.

Restarting the Application

Start Backend

cd backend
python -m uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload

Start Frontend

npm run dev

Understanding the Results

Default Probability

  • 0-33%: LOW risk - Applicant likely to repay
  • 33-66%: MEDIUM risk - Moderate default risk
  • 66-100%: HIGH risk - Significant default risk

Top Risk Factors

The application shows the features that most influenced the prediction:

  • Increases Risk (red): Features pushing the prediction towards default
  • Decreases Risk (green): Features reducing default probability

Each factor includes a human-readable explanation of why it matters.

Troubleshooting

"Backend not available" message

  1. Check if backend is running: curl http://localhost:8000/api/health
  2. If not, restart the backend (see above)

Form not submitting

  1. Ensure all required fields are filled
  2. Check that values are within valid ranges
  3. Open browser DevTools (F12) and check Console for errors

Prediction takes too long

  1. Check backend logs for errors
  2. Ensure model artifacts exist in backend/artifacts/
  3. Try restarting the backend

Need Help?

  • Check PROJECT_STATUS.md for detailed project information
  • Review backend/README.md for backend-specific documentation
  • Check browser console (F12) for frontend errors
  • Check terminal logs for backend errors

Happy Testing! 🎉

Your AI-powered credit risk analyzer is ready to evaluate loan applications with explainable predictions!