Aircraft Maintenance AI Demo & MVP

I need a solo developer who can move quickly from concept to working code. In the first week you will ship a polished visual demo in Glide, Retool, or React that lets engineers and supervisors chat with an AI assistant about EASA Part-145/147 maintenance tasks. AI-powered queries through GPT-4o are the star of this demo, so the chat interface must feel natural and responsive while showcasing checklist validation, image intake (mocked or via a Vision API), and a clean, professional UX.

Immediately after the demo, we transition into a four-to-five-week sprint to turn it into an MVP. I’m leaning toward FastAPI, but Node.js is fine if you are faster there. The goal is a PWA that works offline (IndexedDB), syncs with a PostgreSQL/Supabase backend when connectivity returns, and supports secure digital signatures. Image-based diagnostics can stay mocked for now, yet the architecture should make swapping in a real vision model straightforward.

Deliverables
• Demo PWA (5-7 days): chat UI, GPT-4o integration, basic checklist validation, mocked image upload, slick UX
• MVP backend (4-5 weeks): FastAPI or Node.js services, Supabase schema, offline-first data layer, signature flow, and deployment docs

Budget is €1 600 total, split per milestone above, and I’m ready to start as soon as we align on your proposed stack. Share relevant portfolio links and tell me which tools you’d pick so we can kick off right away.