Larrykj/ecommerce-kenya

AI-powered e-commerce recommendation system for Kenya # 🛒 E-Commerce Recommendation System - Kenya Edition A production-ready, AI-powered e-commerce recommendation system built specifically for the Kenyan market, featuring ML-based personalization, M-Pesa integration, regional insights, and dual-language support. ## 🌟 Key Features ### 🎯 Recommendation Engine - ✅ **User-based Collaborative Filtering** - Find similar users and recommend their preferences - ✅ **Item-based Collaborative Filtering** - Recommend similar products - ✅ **Hybrid Model** - Combines collaborative + content-based filtering using LightFM - ✅ **Trending Items** - Real-time most viewed/sold products - ✅ **Personalized Homepage** - Dynamic user-specific product feeds - ✅ **Context-Aware Suggestions** - Time/location/season-based recommendations - ✅ **Search Auto-Suggestions** - Intelligent search completions - ✅ **Dynamic Discounts & Bundles** - Smart promotional strategies ### 🧠 Machine Learning - **KNN (K-Nearest Neighbors)** - Similarity-based recommendations - **Matrix Factorization (SVD)** - Latent feature extraction - **LightFM** - Hybrid collaborative and content filtering - **Real-time Updates** - Redis caching for instant responses - **A/B Testing** - Continuous recommendation optimization ### 🇰🇪 Kenya-Specific Features - 💳 **M-Pesa Integration** - STK Push payment support - 📍 **Regional Product Trends** - 47 county-based insights - 📶 **Low-Data Optimization** - Optimized for slow connections - 📴 **Offline Capability** - PWA with offline browsing - 🌐 **Dual Language** - Swahili & English support - 🏪 **Local Vendor Promotion** - Priority for Kenyan businesses ## 📁 Project Structure ``` . ├── backend/ # FastAPI Python Backend │ ├── app/ │ │ ├── api/ # API routes │ │ │ ├── endpoints/ # Route handlers │ │ │ └── deps.py # Dependencies │ │ ├── core/ # Core configuration │ │ │ ├── config.py # Settings │ │ │ …