AI Engineer/Developer (LLM & Dataset Build)

We are a private compensation data company. We hold a proprietary, structured database of compensation benchmarking data across firms in the US, UK and Europe. We are building an AI-powered agent that allows clients to query our compensation database in natural language and receive accurate, formatted benchmarking reports in response. What we need you to build: Connect our structured compensation database to an LLM (Claude or GPT-4) via API so it can accurately query and retrieve data in response to natural language questions A clean, branded client-facing interface — login, query input, and formatted report output (PDF or on-screen) Reliable, production-ready performance — not a demo, something we can put in front of paying clients Example of what the finished product does: A client types: "What is the median base salary and carry for an accountant at a European investment bank in London with 6 years of experience?" The agent queries our database, retrieves the relevant data points, and returns a clean benchmarking report with percentile ranges. What we already have: Structured compensation database (ready to connect) Claude/Anthropic API access Clear brief and fast decision-making (non-technical founder, so no internal politics) What we're looking for in you: Demonstrable experience building LLM-powered applications connected to databases (RAG, LangChain, LlamaIndex or similar) Please show us something you've already shipped — a live product or demo that does something similar Ability to work independently with minimal hand-holding Clear communicator — we're non-technical, so we need someone who can explain what they're doing in plain English Project scope and budget: 6-8 weeks Fixed price or day rate considered Budget: open to discussion based on experience and location Potential for ongoing relationship as the product develops To apply: Please send: 1.) A short note on why this project suits you 2.) IMPORTANT: A link or description of a relevant project you've shipped (important) 3.) Your rate/availability. Please note, we are only considering applications from those who have built with LLMs connected to a private database. Please ensure you can demonstrate this before applying.