A RAG-based PDF QA system for local document understanding, combining FAISS vector retrieval with a LangChain pipeline over ChatGLM. Supports chunking, vector indexing, retrieval, and question-answering / summarization on your own PDFs — a practical companion to the retrieval research.