PDF Q&A Agent

This AI agent answers questions from uploaded PDFs using Retrieval-Augmented Generation (RAG).
It demonstrates skills in natural language processing, vector search, and LLM pipelines, allowing users to query documents intelligently.

Features

• Upload any PDF and generate meaningful answers to your questions.
• Uses LangChain for chaining LLMs and handling queries.
• Vector-based retrieval with FAISS for fast and accurate responses.
• Minimal setup required — fully runnable in Google Colab.

Demo Video

Watch the PDF Q&A Agent in action: Loom Video

Demo Screenshots

How it Works

1. PDF is loaded and split into manageable chunks.
2. Embeddings are generated using sentence-transformers.
3. Chunks are stored in a FAISS vector store.
4. Users can input questions, and the model returns precise answers using Retrieval-Augmented Generation (RAG).

Notebook

Run and explore the full Colab notebook here.

Tech Stack

• Python
• Google Colab
• LangChain
• Hugging Face Transformers
• FAISS (Vector Search)
• Sentence-Transformers

Author

AKDGrant