AI RAG Chatbot Developer | LangChain + Pinecone Integration

We’re looking for a developer to create a small but reliable production-ready RAG chatbot connected to our internal knowledge base. Currently, we have internal materials such as playbooks, templates, guides, and reference documents that our team searches manually. The goal is to build a chatbot that can answer questions using this content directly, with strong retrieval accuracy and clear source references, instead of giving generic AI-generated responses. The scope includes: * Ingesting and chunking our internal documents into a vector database, such as Pinecone or pgvector * Building a RAG flow: user query → document retrieval → grounded response with cited sources * Creating a simple, lightweight web-based chat interface that is easy to use and self-contained * Adding a proper fallback when the answer is not available in the knowledge base, to avoid hallucinated responses Preferred technologies include LangChain or LangGraph, Pinecone or pgvector, and Python or Node.js. We’re open to your stack recommendation, as long as the solution is clean, production-grade, and not just a quick prototype. Budget: $500 fixed price We’re also open to splitting the project into milestones, for example: setup and RAG pipeline first, then chat interface and final polish. We’re looking for someone who has real experience building RAG systems, not just calling an LLM API. You should be able to explain your retrieval approach before starting, including your chunking strategy, embedding model choice, and how you would keep answers grounded in the source documents. Please include a short explanation of how you would approach chunking and retrieval for this type of internal knowledge base, along with 1–2 examples of similar RAG projects you’ve worked on.