A fair, transparent credit scoring system for freelancers, gig workers, students, and underbanked communities.
π Live Demo β’ π Documentation β’ π» Tech Stack β’ ποΈ Getting Started
π This is a hackathon demonstration project!
This application uses an in-memory H2 database for quick setup and demo purposes. Your data will be lost when:
- The server restarts
- After periods of inactivity
- When you close your browser
β‘ Do NOT store important or real financial information!
Traditional credit scoring systems are broken and biased:
β Exclude 45% of adults who lack traditional banking history
β Penalize freelancers & gig workers for non-traditional income
β Ignore real financial behavior like on-time rent & utility payments
β Lock out students with no credit cards but responsible spending
β Reinforce inequality by treating lack of history as high risk
Result: Millions of financially responsible people are denied loans, housing, and essential services.
ScoreBridge is an AI-powered financial scoring engine that evaluates real-world money habits, not just credit cards and bank loans.
β
Payment Consistency (35%) - On-time bill payments, rent, utilities
β
Income Reliability (25%) - Steady income patterns, freelance stability
β
Transaction Patterns (20%) - Spending habits, financial responsibility
β
Savings Stability (20%) - Emergency funds, savings growth over time
SBI = 0.35Β·P + 0.25Β·I + 0.20Β·T + 0.20Β·S
Score Range: 300-850 (like FICO, but fairer)
| Feature | Description |
|---|---|
| π€ AI-Powered Scoring | Machine learning model trained to find optimal weights for fair assessment |
| π¨ Beautiful Dashboard | Real-time score visualization with interactive charts and progress bars |
| π Component Breakdown | Transparent view of exactly what impacts your score |
| οΏ½ Smart Recommendations | Personalized tips to improve your financial health |
| π Secure Authentication | JWT-based auth with bcrypt password hashing |
| οΏ½ Fully Responsive | Works seamlessly on desktop, tablet, and mobile |
| β‘ Real-Time Calculation | Instant score results powered by Flask ML service |
| π Demo-Ready | Pre-populated data for easy testing and presentation |
βββββββββββββββββββ ββββββββββββββββββββ βββββββββββββββββββ
β React SPA βββββββΆβ Spring Boot API βββββββΆβ Flask ML API β
β (Port 3000) ββββββββ (Port 8080) ββββββββ (Port 5000) β
β β β β β β
β β’ User Auth β β β’ JWT Auth β β β’ ML Model β
β β’ Dashboard β β β’ Data Storage β β β’ Feature Eng β
β β’ Score View β β β’ API Gateway β β β’ Prediction β
βββββββββββββββββββ ββββββββββββββββββββ βββββββββββββββββββ
β
βΌ
ββββββββββββββββββββ
β H2 Database β
β (In-Memory) β
ββββββββββββββββββββ
- π€ User Module: Registration, authentication, profile management
- π₯ Data Ingestion Module: Financial profile persistence and validation
- π― Scoring Module: AI-powered score calculation and history tracking
- π§ Common Module: Shared utilities, validators, and configurations
- β Java 21 - Modern LTS version with latest features
- π Spring Boot 3.3.5 - REST API, Security, Data JPA
- ποΈ H2 Database - In-memory for hackathon demo
- π Spring Security - JWT authentication with bcrypt
- β Bean Validation - Input validation and sanitization
- βοΈ React 18.2 - Modern UI with hooks
- β‘ Vite 5.0 - Lightning-fast build tool
- π¨ Tailwind CSS 3.3 - Utility-first styling
- π Recharts 2.10 - Beautiful data visualizations
- π» Zustand - Lightweight state management
- π― React Router 6 - Client-side routing
- π Python 3.11 - ML service runtime
- π§ͺ Flask - Lightweight ML API server
- π Scikit-learn - Logistic regression model
- π’ Pandas & NumPy - Data processing
- π― Feature Engineering - Custom financial metrics
- π¦ Maven - Java dependency management
- π¦ npm - Frontend package management
- π³ Docker-ready - Containerization support
Terminal 1: ML Service
cd ml_service
pip install -r requirements.txt
python app.pyβ
ML API: http://localhost:5000
Terminal 2: Spring Boot Backend
# Windows
mvnw.cmd spring-boot:run
# Linux/Mac
./mvnw spring-boot:runβ
Backend API: http://localhost:8080
Terminal 3: React Frontend
cd frontend
npm install
npm run devβ
Web App: http://localhost:3000
Open your browser and go to http://localhost:3000 to start using ScoreBridge!
Email: demo@scorebridge.com
Password: Demo123!
Name: John Doe
Phone: +1234567890
Fill out the financial profile form with:
- π° Accounts: Bank accounts, credit cards, savings
- π Transactions: Income (salary, freelance), expenses (rent, utilities)
- π History: 3-12 months of data recommended
Click "Calculate Score" and get instant results:
- π― SBI Score (300-850)
- π Component Breakdown with visual progress bars
- π Pie Chart showing weight distribution
- π‘ Personalized Recommendations
- View score history on your dashboard
- See latest score and account summary
- Access all past calculations
Try these scenarios for testing:
- β 6+ months on-time payments
- β Steady income from multiple sources
- β Balanced spending patterns
- β Growing savings balance
- β Most payments on time
- β Regular income with minor gaps
- β Reasonable spending habits
β οΈ Modest savings
β οΈ Some late paymentsβ οΈ Irregular income patternsβ οΈ High expense-to-income ratio- β Minimal savings
π¨βπ» Freelancers & Gig Workers - Validate non-traditional income streams
π Students - Build credit history without credit cards
π Renters - Leverage on-time rent payments
π Underbanked Communities - Access fair financial opportunities
πΌ Small Business Owners - Demonstrate cash flow reliability
- Financial Inclusion: Bridge the gap for 45% of adults without credit history
- Fairness: Reward actual financial behavior, not rigid banking rules
- Transparency: Users understand exactly what impacts their score
- Empowerment: Provide actionable recommendations for improvement
- π€ AI scoring model with optimized weights
- π¨ Full-stack React + Spring Boot application
- π Secure JWT authentication system
- π Interactive dashboard with visualizations
- πΎ Data persistence and score history
- π± Responsive mobile-friendly design
- π― Real-time score calculation
- π‘ Personalized recommendations engine
- π Integration with real financial data APIs (Plaid, Stripe)
- π§ Email notifications for score changes
- π PDF score export functionality
- π§ͺ A/B testing for model improvements
- π Multi-language support
- ποΈ PostgreSQL migration for production
- π³ Docker Compose setup
- βΈοΈ Kubernetes deployment configurations
- π Advanced analytics dashboard for admins
This project is licensed under the MIT License - see the LICENSE file for details.
- HackNomics 2025 for the opportunity to build this solution
- Financial inclusion research from the World Bank
- Open-source community for amazing tools and libraries
- All underbanked individuals inspiring us to build fairer systems
- π GitHub: Omar-Mega-Byte/ScoreBridge
- π§ Email: omar.tolis2004@gmail.com
- π¬ Issues: Report a bug or request a feature
β If you find this project helpful, please give it a star!
Made with β€οΈ for a more financially inclusive world