sakthimanocseucet-wq/IOB_KYC ? reverse-engineered prompt

Build me a simple KYC checking app for IOB that can run locally and in Docker.

I want a backend that can receive user details and uploaded images or video, then run checks to help spot fake or suspicious identity documents and face matches. There should also be an AI and machine learning part for the model logic, plus a few basic test scripts so I can verify the models work without doing everything by hand.

Make it easy to start the whole thing on a Windows machine too, since I may use the provided batch and PowerShell scripts. If anything needs current setup details, look up the latest docs online and keep it practical.

The main goal is a working end to end demo that feels like a real KYC flow, with clear error messages, simple responses from the backend, and enough structure that I can improve the detection models later.