Here are the full detailed requirements for the project: **Platform:** * iOS + Apple Watch (watchOS) only * No Android version required **Project Overview:** I want to build a fully functional Apple Watch + iPhone app that can detect scratching/itching behavior using motion data from the Apple Watch. The Watch should handle real-time detection, and the iPhone app should display insights, history, and related health data. **Very Important:** This project must include a **machine learning model** for scratch/itch detection. I do not want just a basic rule-based prototype. A heuristic version can be used only as an early testing step, but the final product must include a proper ML-based detection system. **Apple Watch App Requirements:** * Motion data collection using Apple Watch sensors: * Accelerometer * Gyroscope * Real-time scratch/itch detection on the watch * Haptic feedback when scratching is detected * Notifications/alerts when needed * Efficient battery usage * Sync events and relevant data to the iPhone app * Smooth communication with iPhone using WatchConnectivity **iPhone App Requirements:** * Clean, modern, production-quality UI * Built in SwiftUI * Dashboard showing: * Detected scratching events * Daily / weekly / monthly trends * Event history * User feedback flow so I can confirm whether detections were correct * Data storage and syncing * Settings and controls for notifications and app behavior **Machine Learning Requirements:** * A real ML model must be developed for detecting scratching behavior * The model should use Apple Watch motion data as the main input * The system should be structured so the model can improve over time with more labeled data * Core ML integration is preferred for on-device inference * I want the developer to explain: * how training data will be collected * how the model will be trained * how it will be tested for accuracy * how false positives will be reduced * The final implementation should be designed for real-world usage, not just demo/testing **Health Integration (Apple Health / HealthKit):** The app should integrate with Apple Health and gather/display: * Sleep data * Heart rate / pulse * Temperature, if available from the device/data source * Exercise / activity / fitness data This should work as part of a complete health and wellness tracking experience. **Design / UX Expectations:** * Clean, polished, modern UI * Similar quality level to Apple Health / Apple Fitness type apps * Not a rough prototype * Should feel like a real consumer app that can be published **Important Requirement for Developer:** You must have your **own Apple Watch** for proper development and real-device testing. This project cannot be done properly using simulator only, because the scratch detection depends on live sensor data from the watch. **Performance Requirements:** * Real-time detection should work reliably * ML model should be optimized for watch/iPhone performance * Battery usage should be carefully considered * False positives should be minimized as much as possible * Watch and iPhone syncing should be stable **End Goal:** The final goal is a **fully developed, polished, App Store-ready product** that: * Detects scratching/itching with a machine learning model * Works reliably in real-world conditions * Includes haptic feedback and notifications * Tracks and displays sleep / pulse / temperature / heart rate / exercise / fitness data * Has a clean, modern, fully finished UI * Feels like a complete health/wellness app that can be downloaded from the Apple App Store **Deliverables:** * Full source code * Apple Watch app * iPhone app * Integrated ML model * HealthKit integration * App Store-ready build * Support for testing and bug fixes **What I Need From You:** * Confirmation that you own an Apple Watch * Your proposed ML approach for scratch detection * How you will handle training data and model accuracy * Timeline * Total cost * Examples of relevant Apple Watch / iOS / ML / health apps you have built I do not have final mockups yet, but I want a clean Apple-style health app design. You can also send sample UI references or similar app examples if you have them. Please send me: 1. your total price 2. estimated timeline 3. your ML approach 4. whether you have your own Apple Watch 5. examples of similar work