Fix and Complete an Existing Python RAG Pipeline for PDF-Based Question Answering

I am looking for an experienced Python developer with practical experience in Retrieval-Augmented Generation (RAG), LangChain, and local LLM deployment to help debug, complete, and improve an existing RAG project. The existing code is partially implemented, but some parts of the RAG workflow need debugging, improvements, and proper integration. What I Need: The main goal is to adapt and complete the existing pipeline so that it can work with user-provided PDF documents. The expected functionality includes: - Load and extract text from selected PDF documents. - Generate embeddings and store them in the vector database. - Retrieve relevant passages based on a user's question. - Generate answers using the existing LLM and retrieved document context. - Identify and fix bugs or integration issues in the current workflow. - Ensure the pipeline can run successfully from document ingestion to answer generation. Expected Deliverables: - A working and debugged version of the existing Python code - Functional PDF ingestion and document-based question answering - Brief instructions explaining how to run the updated project Important Notes: This is an existing project, not a request to build a RAG system from scratch. I will provide the current source code.