A Retrieval-Augmented Generation (RAG) based Ecommerce Customer Support Chatbot built with LangChain, OpenAI GPT, FAISS, and Gradio. This bot retrieves answers from a structured FAQ dataset and generates helpful responses for customers.
Retrieval-Augmented Generation (RAG) for accurate, source-based answers
Semantic Search powered by OpenAI embeddings & FAISS
Interactive Web UI using Gradio (runs in Google Colab or locally)
Easy to adapt with your own FAQ or knowledge base
Load your FAQ dataset (JSON, TXT, PDF, etc.)
Split into chunks for better retrieval
Embed with text-embedding-3-small model
Store in FAISS vector database
Retrieve relevant chunks when a customer asks a question
Generate answer using GPT-4/4o with the retrieved context