← Back to Projects E-commerce 
Hybrid Recommendation Engine
Production-ready recommendation system combining collaborative filtering (Surprise) and content-based filtering with cold-start handling, Redis caching, and A/B testing.
Overview
Production recommendation system combining multiple filtering approaches with A/B testing infrastructure.
Architecture
- Matrix factorization via Surprise library
- TF-IDF content-based filtering
- Hybrid combiner with configurable weights
- Redis caching for low-latency serving
- FastAPI with A/B testing endpoint
Key Features
- Cold-start handling for new users/items
- Redis-backed response caching
- A/B testing for model comparison
- Real-time recommendation updates