ChicaneAI

AI-powered Formula 1 analytics and race prediction web app.

Features

• Race predictions - win probability for each driver using a Gradient Boosting model trained on FastF1 data
• H2H comparisons - head-to-head driver stats pulled live from FastF1 with an ML prediction card
• History - past predictions verified against real results
• Season calendar - 2026 F1 race schedule

Track Record

Accurately predicted Antonelli (Mercedes) as the 2026 Miami GP winner.

Stack

Frontend

• React + Tailwind CSS (Vite)

Backend

• FastAPI (Python)

ML

• scikit-learn (Gradient Boosting + Random Forest)
• FastF1 (real F1 timing and results data)

Documentation

• docs/CONTEXT.md - project context and current architecture notes
• docs/design.md - design system reference
• docs/security_audit.md - latest security audit notes

Setup

Frontend

cd frontend
npm install
npm run dev

Runs on http://localhost:5173

Backend

Recommended: Python 3.12.

cd backend
Remove-Item -Recurse -Force .\venv -ErrorAction SilentlyContinue
python -m venv venv
.\venv\Scripts\Activate.ps1
python -m pip install --upgrade pip
python -m pip install -r requirements.txt
python -m uvicorn app.main:app --reload

Runs on http://127.0.0.1:8000

Health check: http://127.0.0.1:8000/api/health

Run backend tests:

cd backend
.\venv\Scripts\Activate.ps1
python -m pip install -r requirements.txt
python -m pytest tests

For the contact form, copy backend/.env.example to backend/.env and fill in the Resend/contact values. The H2H cache TTL can be adjusted with H2H_CACHE_TTL_SECONDS.

If PowerShell blocks the activation script, allow local scripts for your user:

Set-ExecutionPolicy -Scope CurrentUser RemoteSigned

Then run:

.\venv\Scripts\Activate.ps1

What's Next

• Model improvements
• Deployment

License

MIT