RAG System with Gemini & MongoDB

This project explores the Retrieval-Augmented Generation (RAG) architecture. It uses Google's generative AI to understand user questions and retrieves relevant information from a knowledge base stored as vector embeddings in MongoDB Atlas. The goal was to build a system that provides accurate answers based on a specific set of documents.

• PDF document processing and text chunking
• Text embedding using the Gemini API
• Vector storage and search with MongoDB Atlas
• Generative question-answering based on retrieved context