AI Developer Needed — Build Local RAG-Based Espresso Troubleshooting Assistant

Description: I'm looking for a developer to build a small AI-powered assistant for home baristas. A user describes a problem with their espresso shot (sour, bitter, watery, pulling too fast or too slow), and the assistant diagnoses the likely cause and recommends a specific fix — grounded in real brewing knowledge, not a generic AI guess. This is a well-scoped, small first version — not a full product build. I'd rather have something working, clean, and clearly explained than something over-engineered. What's in scope: User inputs a problem in plain language (e.g. "shot tastes sour and pulled in 15 seconds"), optionally with grind size, dose, and yield The assistant retrieves relevant guidance from a small set of real brewing documents (5–8 sources) Uses a RAG (retrieval-augmented generation) approach with an LLM API (e.g. Anthropic Claude or OpenAI) to generate the diagnosis Returns a clear answer: likely cause, the fix to try, and which source it's based on A simple interface: enter a problem, get a diagnosis What I'll provide: The source documents for the knowledge base (brewing guides, troubleshooting references) What I need from you: Build the agent to run locally on my machine (no cloud deployment needed for this phase) Hand over the complete, well-organized codebase at the end One walkthrough call after handover explaining how the code works, so I can maintain and update it myself going forward