What you can look at
Any measure, the windows that matter
24 hours, 7 days, 30 days, 3 months, 6 months, a year. Your usual range as a band, touch-and-drag inspection, and gaps drawn as gaps.
The long game
Half a year a week at a time, for change too slow to feel.
Body composition
Weight, fat, lean and muscle mass — one source per measure, never averaged across devices.
Segmental body
Per-limb fat and muscle from a body scan, including the trunk-to-legs ratio.
VO₂ max
Every method side by side, device reading and Brin’s own estimate, shown as a range rather than averaged into one invented figure.
Training
Runs compared with runs, each lift against its own history: load, volume, top set, estimated 1RM, elevation as context.
Muscle load
What the week actually trained, split by muscle group.
Healthspan
Eight domains tracked separately — aerobic, cardiovascular, metabolic, body composition, sleep, strength, biomarkers, daily movement. Never blended into a “body age”.
Progress
Week over week and month over month, with the comparison invalidated when the measurement method changes.
One source per measure, per day
Brin models 18 data sources across 68 canonical metrics — 16 platforms, plus laboratory results and anything you enter by hand. When three devices report the same session it resolves them into one, keeping its real name — and it will not average two devices into a number neither of them reported. A provider’s missing-value marker is not a measurement: a −1 ms HRV is not a low reading, it is not a reading.
| Live | Modelled |
|---|---|
| Apple Health, WHOOP, Withings | Oura, Garmin, Polar, Coros, Suunto, Concept2, Wahoo, Strava, TrainingPeaks, Google Health API (Fitbit and Pixel Watch), Health Connect, Dexcom, Samsung Health, laboratory results, manual entry |