Leverages Retrieval-Augmented Generation (RAG) to answer natural language questions based on the content of uploaded documents (PDF, DOCX, TXT). Cites sources within the documents.