AI/ML Engineer to build an AI powered clinical document assistant.

We’re looking for an AI/ML engineer to help us build an MVP for a clinical document assistant. The idea is simple: clinicians upload patient documents such as discharge summaries, laboratory reports, referral letters, and clinical notes. The system should extract the important information, organize it, and allow the clinician to ask questions about the documents using natural language. The challenge is making this reliable. We don't want a chatbot that simply generates plausible answers. Responses should be grounded in the uploaded documents and provide references to the source information. What we need * Document upload and processing pipeline * PDF/text extraction * LLM-powered information extraction * RAG-based question answering * Vector database integration * FastAPI backend * Secure user/document isolation * REST APIs for frontend integration * Basic web interface for testing the MVP * Cloud deployment Expected MVP A user should be able to: 1. Create an account 2. Upload a clinical document 3. Wait for it to be processed 4. Ask questions about the document 5. Receive an answer grounded in the document 6. See which part of the document supports the answer We're open to using OpenAI, Anthropic, Gemini, or other reliable models. The architecture should make it possible to change providers later without rebuilding the application. What I'm looking for I care more about someone who can think through the problem than someone who simply follows a specification. The code should be clean, documented, and easy for another developer to take over. I also want someone who will tell me when an approach isn't technically sound rather than just implementing it because I asked for it. If the MVP goes well, there will be additional work around document classification, structured clinical data extraction, and workflow automation. Please include: * 2–3 relevant projects you've built * Your proposed approach for the MVP * Your availability * Your hourly rate I'm looking for someone interested in building this beyond the initial MVP, not just completing a one-off task.