ros4n/DocumentAI ? reverse-engineered prompt

Build me a web app for scanning documents and filling forms with OCR and AI help.

I want users to sign up, log in, and keep a saved profile with contact, job, ID, and custom details that can be reused later. They should be able to upload a photo or PDF, or take a picture from their camera, then extract the text, review it, and optionally auto fill forms from their saved profile using an OpenAI compatible API if they turn that on. Add a scan history page where people can search, filter, rename, delete, and compare the original scan with the filled version side by side. Include basic image tools like crop, zoom, and a fullscreen before and after view. Make it feel like a modern installable app that works well on mobile too, and set it up so the frontend and backend can run locally with a simple dev setup. If you need to check current docs for any libraries, go ahead and look them up online.