Fairly AI is a high-performance, production-ready SaaS platform designed to detect, analyze, and mitigate algorithmic bias in machine learning models. Built with a premium Stripe-inspired UI and a robust FastAPI backend.
- 📊 Advanced Fairness Metrics: Real-time auditing for Statistical Parity, Equal Opportunity, and Disparate Impact.
- ⚡ Zero-Latency Simulations: High-speed "What-If" simulation engine using in-memory model caching.
- 🎨 Premium SaaS UI: Interactive dashboard with glassmorphism, smooth animations, and high-fidelity Recharts.
- 🛡️ Bias Mitigation: Automated remediation strategies using industry-standard reweighting and resampling.
- 📄 Professional Reports: One-click PDF export for compliance and audit logging.
- FastAPI: Asynchronous high-performance API.
- Scikit-Learn: Core ML and evaluation metrics.
- Joblib: Efficient model persistence.
- Pandas/Numpy: High-speed data processing.
- Tailwind CSS: Modern, responsive styling.
- Framer Motion: Smooth interactive animations.
- Recharts: Data visualization.
- Lucide React: Premium iconography.
- Python 3.10+
- Node.js 18+
cd backend
python -m venv venv
source venv/bin/activate # On Windows use: venv\Scripts\activate
pip install -r requirements.txt
uvicorn main:app --reloadcd frontend
npm install
npm run dev├── backend/
│ ├── routes/ # API Endpoints (Analyzer, Predict, Chat)
│ ├── services/ # Logic (Bias Detection, Mitigation)
│ ├── main.py # FastAPI Entry Point
│ └── .env # Secrets (Ignored by Git)
├── frontend/
│ ├── src/
│ │ ├── components/ # Reusable UI Components
│ │ ├── pages/ # Dashboard and Analytics
│ │ └── services/ # API Interaction
│ └── tailwind.config.js
└── README.md
Fairly is an open-source project aimed at making AI more equitable. Feel free to open an issue or submit a pull request! 🚀
MIT License. Built with ❤️ for the Google Solution Challenge.