This project demonstrates the development of an AI-powered chatbot capable of answering user queries based on the content of uploaded PDF documents. Built using Retrieval-Augmented Generation (RAG), it combines LangChain, FAISS, and HuggingFace Transformers to retrieve and generate contextually accurate responses.
Users can upload one or more PDF files, ask natural language questions, and receive intelligent answers extracted from the document contents. The project showcases practical applications of agentic AI, embeddings-based retrieval, and open-source LLMs to enable domain-specific conversational intelligence.
Key components include:
LangChain for chaining retrieval and generation logic
HuggingFace sentence-transformers (MiniLM) for text embeddings
FAISS for efficient vector similarity search
Flan-T5 model for text generation
Streamlit for an interactive user interface
This project lays the foundation for document-grounded conversational agents that can power enterprise knowledge assistants, research bots, or intelligent PDF readersโall running locally without API dependencies.