Phoebe

The app continuously gathers sleep duration, heart‑rate variability and resting heart rate from an Apple Watch via HealthKit, and combines these biometrics with hourly local atmospheric data—barometric pressure, humidity and temperature—sourced from WeatherKit. A machine‑learning model blends these 27 inputs into a personalized risk score that updates over time, producing both immediate alerts and a seven‑day forecast of migraine likelihood.

When the model predicts a heightened risk, the user receives a contextual notification suggesting actions such as adjusting sleep, schedule or medication. The system also supports voice‑activated logging, allowing users to record migraine severity, location and medication without manual input. An on‑device AI health assistant can answer queries about individual triggers and patterns using only the stored diary data.

Targeted at individuals who experience migraines, the utility aims to provide proactive, data‑driven warnings based on personal biometrics and micro‑climate conditions, helping users anticipate attacks before symptoms appear.