I need a proof-of-concept Windows desktop app that can take over a Nikon DSLR connected by USB, trigger photo or short-video capture, pipe the footage through an AI background-replacement service, then present the finished result in a full-screen, touchscreen-friendly interface. Core flow 1. Control a Nikon body using the official SDK (or a reliable open-source wrapper). 2. Capture single images and up to 30-second clips. 3. Immediately send the raw file to an AI endpoint. I have access to both OpenAI and Stable Diffusion APIs and I’m open to your recommendation on which is faster and more cost-effective for real-time background/scene generation. 4. Show the composited output in a simple gallery view where users can swipe through, accept or discard. 5. On acceptance, allow export to a local folder and auto-generate a QR code that points to the saved file for easy mobile download. Tech stack is flexible: Python (PyQt, Kivy), C# (.NET/WPF), or Electron—choose whatever gets us to a stable MVP quickest, as long as setup on Windows is straightforward. As part of the delivery I’ll need: • Well-commented source code and build/run instructions • A brief README describing any third-party libraries, API keys, or SDK installers required • A short screen-capture video showing the full capture-to-export sequence If something in the scope feels unclear, flag it early so we can keep the prototype tight and focused.