Scalable Recommendation API ⚡

High-Performance Microservice A lightweight, low-latency recommendation engine designed to serve personalized content with sub-50ms response times. Unlike standard ML notebooks, this project focuses on the serving layer, optimizing for throughput and API design.

🏗️ Architecture

• Framework: Flask (Python)
• Caching: In-Memory LRU Cache to minimize compute for frequent queries.
• Algorithm: Content-Based Filtering (Cosine Similarity).

🚀 Performance

• Average Latency: ~35ms
• Throughput: Handles 500+ requests/sec on local testing.

🔌 API Endpoints

• GET /health: System status check.
• POST /recommend: Input user_id, returns JSON list of product_ids.