Hi Community,
I'm a developer considering building a health & wellness iOS app and want to verify compliance with Apple's guidelines before investing development time.
BACKGROUND: I want to create an app that helps users understand their health patterns by analyzing various data inputs including screen time and health metrics from HealthKit.
PLANNED APPROACH: The app would use a hybrid ML strategy:
- NEW USERS: Get insights from a generalized ML model trained on anonymized data from existing users (solves cold-start problem)
- AFTER ~30 DAYS: Transition to personalized on-device model trained on the user's own data for better accuracy
DATA HANDLING:
- Users grant DeviceActivity & HealthKit permissions explicitly
- Users select which app categories they want to include
- Optional opt-in to share anonymized data for improving the generalized model
- Transition to fully on-device analysis after sufficient data
QUESTIONS:
- Are there specific restrictions on using DeviceActivity data and sending them to a backend for ML training ?
- Are there specific restrictions on using HealthKit data and sending them to a backend for ML training ?
Any guidance would be greatly appreciated before I proceed!
Thanks!