ImpactLens

Full-stack business impact analysis tool with a React/Vite frontend, an Express backend, and a FastAPI ML service. The frontend now calls the backend /predict API, which in turn delegates to the ML service.

Stack

• Frontend: React 18 + Vite (located in frontend/, source in /src)
• Backend: Express (port 8000) in backend/
• ML service: FastAPI (port 9000) in ml/prediction_service

Quick Start (Docker Compose)

docker-compose up --build

Services and ports:

• Frontend: http://localhost:5173
• Backend API: http://localhost:8000
• ML service: http://localhost:9000

Volumes keep source code mounted for hot reload:

• Frontend mounts the repo to /app so /src is available to Vite
• Backend mounts backend/public and backend/datasets

Manual Development (no Docker)

1. ML service

cd ml/prediction_service
python -m venv .venv && .venv/Scripts/activate  # Windows
pip install -r requirements.txt
uvicorn app:app --host 0.0.0.0 --port 9000

2. Backend

cd backend
npm install
set ML_SERVICE_URL=http://localhost:9000
npm start

3. Frontend

cd frontend
npm install
set VITE_API_URL=http://localhost:8000
npm run dev -- --host --port 5173

API Reference

• POST /predict
  • Body: { businessType, scale, locationKey, locationLabel?, contextSignals?, query? }
  • Returns prediction payload + AI explanation
• GET /predict/locations – available location profiles
• POST /simulate – simulation endpoint
• GET /health – service health check

Configuration

• Frontend: VITE_API_URL (defaults to http://localhost:8000)
• Backend: ML_SERVICE_URL (defaults to http://localhost:9000)

Data Flow

1. User adds a business and places it on the map
2. Frontend sends a request to backend /predict
3. Backend calls ML service /predict and generates an AI explanation
4. Frontend merges ML results with local analytics for display and exports

Testing / Validation

• Hit http://localhost:8000/health to confirm backend is running
• From the frontend, ensure predictions render after placing a business on the map