# Product Recommender with Ecommerce Store

> A RAG-powered chatbot that delivers personalized product recommendations, integrated with an eCommerce store to allow users to conveniently purchase suggested products.

- **Category:** AI/ML + E-commerce
- **Scope:** AI/Recommendation + E-commerce
- **Client:** In-house Project
- **Year:** 2024
- **Duration:** 8 months
- **Page:** https://www.cognidiv.com/our-work/product-recommender-ecommerce-store

## Overview
At the core of this project is a RAG-powered recommendation system designed to provide accurate and personalized product guidance. The system extracts and processes data from multiple sources, structures it into a knowledge base, and generates vector embeddings for efficient retrieval. When a user interacts with the chatbot, their query is transformed into vectors and matched against the knowledge base. By re-ranking results and applying contextual filtering, the chatbot provides precise, reliable recommendations aligned with the user’s goal. Continuous feedback loops enable the model to improve over time, ensuring that responses remain relevant and trustworthy. The platform integrates a fully functional ecommerce store, allowing users to browse, filter products, and securely checkout with multiple payment options. This delivers both intelligent recommendations and real-world purchasing convenience.

## Key features
- RAG-powered personalized recommendations
- Product catalog browsing and filtering
- Secure checkout with multiple payment options
- Continuous learning via feedback loop

## Challenges solved
- Building scalable RAG pipelines
- Efficient vector storage and retrieval
- Ensuring product relevance in recommendations
- Integration of AI with eCommerce flow

## Technologies
Next.js, Tailwind CSS, FastAPI, Node.js, MongoDB Atlas, Python, MongoDB Atlas VectorDB, RAG, Google Generative Ai, Sentence Transformers, Oauth

---
Source: https://www.cognidiv.com/our-work/product-recommender-ecommerce-store