Project Bolt is an AI-powered recommendation engine designed to analyze user ratings of movies and TV series to generate personalized viewing suggestions. It combines deep learning models with real-time API integrations for accurate and engaging recommendations.
- 📊 User Profiling: Automatically detects genres, years, and types the user prefers.
- 🧠 Neural Collaborative Filtering (NCF): Advanced model for personalized recommendations.
- 📦 TMDB API Integration: Enriches content with genres, titles, years, and ratings.
- 📺 React + Tailwind Frontend: Clean and fast UI to display recommendations.
- 🌐 API-ready backend for real-time inference.
| Layer | Tech Stack |
|---|---|
| Frontend | React, TypeScript, TailwindCSS |
| Backend API | Python, FastAPI or Flask |
| ML Model | TensorFlow/Keras (NCF) |
| Data Source | TMDB API (https://www.themoviedb.org/) |
project/
├── recommendation_system/
│ ├── train_model.py # Trains NCF model
│ ├── ncf_model.py # Defines model architecture
│ ├── recommendation_server.py # API backend (Flask/FastAPI)
│ ├── NCF_Training_Notebook.ipynb
├── src/ # React Frontend
│ ├── App.tsx
│ ├── hooks/ # Custom hooks for profile + recommendations
│ ├── services/ # TMDB API service
├── index.html
├── package.jsongit clone https://github.com/yourusername/project-bolt.git
cd project-boltcd project
npm installcd project/recommendation_system
pip install -r requirements.txtCreate a .env file with your TMDB credentials:
VITE_TMDB_API_KEY=your_api_key
VITE_TMDB_ACCESS_TOKEN=your_access_tokenConvert your ratings into the correct format and run:
python train_model.py --data ratings.json --output my_model.h5python recommendation_server.pynpm run devPRs and suggestions are welcome. Let's build smarter recommendation systems together!
MIT © Solo.Han