I want to launch a straightforward web tool that lets visitors upload a clear photo of their face and instantly learn whether visible acne is present. The only artificial-intelligence feature I need right now is image-recognition that can correctly detect acne; no predictive analytics or recommendation systems are required at this stage, though the overall architecture should leave room for adding those later if we choose. The flow is simple: the user drops an image, your backend runs a trained model, and the front end returns a visual highlight of the acne areas plus a concise “acne detected / not detected” result—optionally accompanied by a basic severity score. Consistency across different lighting conditions and skin tones is important. I’m open on the exact tech stack as long as you stay with mainstream tools (e.g., Python with TensorFlow or PyTorch for the model, OpenCV for pre-processing, React or Vue for the UI). Hosting can be on AWS, GCP, or any similarly reliable cloud service. Deliverables • Trained and documented acne-detection model • Responsive web interface with upload, processing status, and results view • Deployed demo link plus install/run instructions for local and cloud environments • Brief technical report covering data used, training process, and evaluation metrics (precision/recall on a held-out test set of at least 100 images) Acceptance criteria • ≥ 90 % precision and recall on the provided test set • Full end-to-end workflow completes in under 10 seconds per image on standard cloud hardware • Clean, modern UI that works on desktop and mobile Provide a realistic timeline and note any data requirements you may have so I can secure images or annotate them if needed.