Built a multi-agent RAG platform that enables contextual Q&A across large PDF sets. Integrated LangChain for orchestration, Pinecone for vector storage, and OpenAI for response generation. Supports document ingestion, chunking, embedding, and retrieval pipelines.