It starts with your Apple Watch.
With your permission, Unfurl reads five categories from Apple Health. The core model uses data recorded by Apple Watch, rather than treating every source in Health as equivalent.
- Heart rate
- Your recent pulse provides a signal of how hard your body is working.
- Resting heart rate
- A personal baseline gives that pulse context. The same heart rate can mean something different for different people.
- Steps
- Movement helps the model distinguish physical activity from a quieter reading.
- Sleep
- Recorded sleep helps estimate overnight recovery. When a sleep record has not arrived yet, the model may provisionally infer sleep from supported quiet periods.
- Workouts
- Workouts recorded by Apple’s Workout app help account for exercise and the recovery period afterward. Third-party workout apps are not currently supported by this model.
Stress Level: a signal of physiological strain.
The model compares heart rate with your resting baseline and puts it in context with movement, sleep and workouts. The result is an estimate from 0 to 100.
The app groups it into Rest (0–25), Low (26–50), Medium (51–75) and High (76–100). It does not use beat-to-beat heart rate variability (HRV).
During movement and workouts, a number may be unavailable. After a workout, it returns when a suitable sedentary reading arrives, subject to the model’s recovery rules. Furl stays in recovery mode for the first hour after a supported workout.
A high number cannot tell you why it is high. It does not measure your thoughts, diagnose anxiety or tell you how you should feel.
Body Energy: your reserve, estimated.
Body Energy runs from 5 to 100. The model looks back over recent days and estimates how sleep, rest, strain, movement and workouts add to or draw from your reserve. It is not a direct measurement of energy, calories or fitness.
Today shows what contributed to the change, along with how much the estimate charged or drained. Week and month views help you look for patterns over time.
When a long stretch of watch data is missing, the model can estimate the gap using your usual patterns after readings resume. It labels that contribution “Watch off (estimated)”. That estimate is not recorded sleep and cannot tell how long you actually slept.
A day with Unfurl.
Check your estimated overnight recharge in the morning, glance at your stress and energy during the day, and look back at the factors in the evening. Widgets and watch complications put the latest available result close at hand.
The iPhone and Apple Watch calculate locally and share results. Updates depend on new Health data, device connectivity and the background time Apple makes available. They are periodic, not a continuous medical monitor. Older readings are shown with their appropriate state.
Optional notifications can offer a morning summary or a timely prompt. Their delivery is limited so Furl has space to be helpful without constantly interrupting.
A useful estimate, with honest limits.
- The model is new and was initially calibrated using one person’s data. Results may read higher or lower for you.
- Body Energy needs about 36 hours of wear before its initial calibration is complete. Historical data availability affects the first days shown.
- Missing data, loose fit and delayed Health synchronization affect the estimates.
- Watch-off periods are estimated from typical patterns, not measured sleep.
- Furl’s messages follow rules based on the estimates. They are not personalized clinical advice.
Use the numbers alongside how you feel. Unfurl is a wellness app, not a medical device, and is not intended to diagnose, treat or prevent a condition. Ask a qualified health professional before making medical decisions.