Back to projects
Machine Learning · FinTech

Loan Evaluation & Risk Assessment System

AI-powered loan approval prediction and risk assessment platform.

Enter applicant details below and the evaluation engine — the same risk, credit, geographic and recommendation logic as the original project — returns a live decision.

PythonStreamlitPandasNumPyPlotlyScikit-learnPydanticREST APIReactTypeScript
Live demo

Loan application

All fields are validated on the client and re-validated on the server before the evaluation models run.

1

Personal Information

18–100

2

Financial Information

Debt-to-income ratio drives financial risk

3

Loan Information

Minimum 1,000

4

Credit Information

300–850

5

Location

Regional economic risk model

About the project

This project demonstrates an AI-powered loan evaluation workflow that analyses an applicant's financial, credit, loan and geographic information to produce a risk assessment and a loan recommendation.

The original system is a Python/Streamlit application built from modular components — RiskAnalyzer, CreditScorer, GeolocationAnalyzer, LoanRecommender, DataProcessor and InputValidator. For this portfolio the Streamlit UI is separated from the business logic, which is exposed through a JSON API at POST /api/public/evaluate-loan and consumed by this page.

Every number on the results dashboard — risk score, credit grade, interest rate, amortised monthly payment, approval probability — comes from the evaluation engine on the server, using the same thresholds, weights and formulas as the repository.

Evaluation pipeline

  1. 1Applicant form (React)
  2. 2Client-side schema validation
  3. 3POST /api/public/evaluate-loan
  4. 4Server-side re-validation
  5. 5Risk · Credit · Geo · Recommender models
  6. 6Structured JSON response
  7. 7Results dashboard

AI-powered risk assessment

Weighted financial, credit and employment risk scoring.

Credit score analysis

Score adjustment by defaults and credit-history length, with grading.

Geographic risk assessment

State-level unemployment, income and crime risk model.

Loan recommendations

Affordability caps, credit-tier pricing and amortised payments.

Interactive evaluation

Full applicant form evaluated live through the API.

Risk classification

LOW / MEDIUM / HIGH categories with contributing factors.

Financial analysis

Monthly payment, total cost and alternative loan structures.

Analytics dashboard

Session-level approval rate, average risk and credit metrics.