I’m building a native-feeling iOS and Android app that lets users upload a photo and, in a single tap, watch it come back beautifully color-graded. The core of the project is an on-device (or extremely lightweight cloud) model that reads the image’s existing tone, lighting, and overall mood, then applies a cinematic grade that looks fresh while keeping skin tones and fine details intact. The AI engine must reliably handle: • Tone adjustment • Lighting correction • Detail preservation A single-photo workflow is perfectly fine for version 1, but designing the pipeline so it can easily scale to batch processing later would be a welcome plus. I need the complete mobile application: model integration (TensorFlow Lite, Core ML, or a comparable solution), a responsive UI that previews changes in near-real time, and export back to the camera roll without noticeable quality loss. Clean, well-commented code and a short README explaining build steps will be the measure of completion.