An enterprise-grade peer-to-peer lending platform combining a Node.js REST API, Python ML credit-risk services, and Ethereum smart contracts.

code/
├── backend/            # Node.js / Express REST API
├── blockchain/         # Solidity smart contracts (Hardhat + Truffle)
├── ml_services/        # Python ML models & compliance framework
└── README.md           # This file

• Node.js ≥ 18, npm ≥ 9
• Python ≥ 3.11, pip
• Docker & Docker Compose (for containerised setup)
• MongoDB 7 & Redis 7 (or use Docker)

git clone https://github.com/quantsingularity/LendSmart
cd LendSmart/code
cp backend/.env.example backend/.env

Each service ships its own Dockerfile. Use the docker-compose.yml provided at the repo root (outside this code/ directory) to spin up the full stack:

docker compose up --build

Services started:

Backend

cd backend
npm install
npm run dev        # Development (nodemon + debug port 9229)
npm start          # Production
npm test           # Jest test suite

ML Services

pip install -r ml_services/requirements.txt
# Train the model (first time)
python -m ml_services.credit_risk.train_model
# Start the prediction server
python -m ml_services.credit_risk.prediction_service --server --port 8000

Blockchain

cd blockchain
npm install
npx hardhat compile
npx hardhat test
npx hardhat node                              # Local chain
npx hardhat run scripts/deploy.js --network localhost

Node.js + Express REST API with MongoDB, Redis, JWT auth, rate limiting, audit logging and GDPR compliance.

backend/src/
├── controllers/    authController, loanController, userController, adminController
├── models/         User.js, Loan.js  (Mongoose schemas)
├── routes/         authRoutes, loanRoutes, userRoutes, adminRoutes
├── services/
│   ├── ai/                 aiService.js       – calls ML prediction endpoints
│   ├── blockchain/         blockchainService.js – ethers.js multi-network client
│   ├── compliance/         complianceService.js
│   └── notification/       notificationService.js
├── middleware/
│   ├── auth.js             JWT protect / authorize helpers
│   ├── monitoring/         errorHandler, healthCheck, metricsCollector
│   ├── security/           authMiddleware, rateLimiter, securityMiddleware
│   └── validation/         inputValidator (Joi schemas)
├── compliance/     auditLogger.js, gdprCompliance.js
├── config/         database.js, redis.js, blockchain.js, encryption.js
├── security/       authService.js
├── gateway/        apiGateway.js
└── utils/          logger.js, complianceMonitor.js

Key environment variables → see backend/.env.example.

Smart contracts for on-chain loan lifecycle management.

• contracts/ – Hardhat contracts (LendSmartLoan.sol, LoanContract.sol, Lock.sol, MockERC20.sol)
• truffle/ – Legacy Truffle contracts (LoanManager.sol, BorrowerContract.sol)
• test/ – Hardhat Mocha/Chai test suite
• scripts/ – Deployment scripts

See blockchain/README.md for full toolchain instructions.

Python ML pipeline for credit scoring, risk assessment, alternative data scoring, and regulatory compliance.

• credit_risk/ – Ensemble ML models (sklearn, XGBoost, LightGBM, SHAP), Flask prediction API
• compliance/ – Fair-lending, AML, GDPR, adverse-action notice generation
• integration/ – LendingSystem orchestrator combining all ML and compliance components

See ml_services/README.md for full service documentation.

# Backend
cd backend && npm test

# ML Services
pytest ml_services/credit_risk/tests/ -v

# Smart Contracts
cd blockchain && npx hardhat test