Designed and developed a Retrieval-Augmented Generation (RAG) chatbot capable of answering questions based on PDF documents. Integrated LangChain, FAISS vector database, and Hugging Face models to enable context-aware document search and response generation. Implemented a user-friendly Streamlit interface supporting real-time PDF uploads, automatic chunking, and semantic vector embedding. The system uses Ollama's Gemma3:1B LLM for generating accurate and efficient responses. Optimized for CPU usage to ensure smooth performance even in low-resource environments. Ideal for internal documentation Q&A, academic reference bots, or any use case involving interactive document comprehension.