Know why your energy is what it is.
LockIn90 ingests your wearables, health, CGM, calendar and weather, runs them through an explainable scoring pipeline, and tells you the why: your afternoon dip is yesterday's late caffeine plus poor sleep.
Raw signals in, the why out.
Ingest — Connect your data
Link Oura, WHOOP, Garmin, Polar, Apple Health, Dexcom CGM, Google Calendar and weather — plus your logged sleep, workouts, nutrition, substances and mood. One unified timeline.
Model — We model the effects
Each event runs through effect generators with PK half-life models — a 4pm espresso's caffeine decays on a real curve, sleep debt rebounds — then calibration tunes the magnitudes to you.
Score — Energy curve + score
The pipeline produces a daily 0–100 energy score, a 34-slot circadian curve that maps your real peaks and dips, and a 31-day forecast of where you're headed.
Act — Act on the why
Plain-language insights tell you the cause and the fix — "move your caffeine cutoff 3h earlier." Time hard work to your peaks, protect recovery windows, break the loops.
Everything is explainable.
No black box. Every number traces back to a cause you can see, question and act on.
Explainable energy score
One 0–100 number plus a 34-slot circadian curve mapping your real high and low windows — never a population average. Plan your day around your real peaks.
31-day forecast
Project planned events, habits and recovery forward to see where your energy is heading — and head off a crash before it lands. Schedule the hard weeks with eyes open.
"Why this score?" attribution
We show our work. Every point up or down is traced to a driver: late caffeine −9, short sleep −7, morning lift +6. Fix the actual cause, not a guess.
Pattern detection & insights
Statistically-tested loops surface as plain-language insights — "afternoon dips follow late espresso 4 days in 5, p<0.05." Break repeating energy traps.
Coach + E2EE messaging
Share a privacy-gated energy view with your coach and message end-to-end encrypted. They see trends, never your raw private logs. Real coaching, real privacy.
Calibration that learns you
Over ~30 days the model tunes effect magnitudes to your physiology — your baselines, your thresholds — and reports its confidence. Sharper attribution the longer you use it.
Auto wearable sync
Background sync pulls sleep, HRV, resting heart rate, strain and readiness from every connected source, with historical backfill. Set it once, it just stays current.
3-tap logging
Coffee, training, a hard meeting, a bad night — logged in seconds. CGM and calendar fill in the rest automatically. Capture context without journaling.
MODERATE · −22 vs your baseline · Energy 62
Your afternoon dip wasn't a mystery.
It was yesterday's 4pm espresso decaying on its half-life, plus six hours of sleep. LockIn90 quantifies each driver so the fix is obvious.
Morning training: +6.
Late caffeine (4pm, t½ ~5h): −9.
Short sleep (6h 04m): −7.
Bring every signal.
Wearables, health platforms, continuous glucose, your calendar and the weather outside — all on one timeline, all explainable.
Oura
Sleep stages, readiness, HRV & resting heart rate from your ring. RING · SLEEP · Connected.
WHOOP
Strain, recovery & continuous HRV from the band. BAND · STRAIN · Connected.
Garmin
Body Battery, stress, sleep & activity from your watch. WATCH · BATTERY.
Polar
Nightly recovery, training load & orthostatic HRV. WATCH · HRV.
Apple Health
Aggregated sleep, HR, workouts & mindful minutes. PLATFORM · Connected.
Dexcom CGM
Continuous glucose feeds the glycemic effect generator. CGM · GLUCOSE.
Google Calendar
Meetings & travel become demand load on your curve. SCHEDULE.
Weather & air
Heat, pressure, pollen & air quality as external drivers. ENVIRONMENT · Auto.
Clinical-grade, not vibes.
Effect generators with pharmacokinetic models, calibration that reports its confidence, and patterns that only surface when they pass a significance test.
Half-life curves, not flat penalties
A substance doesn't hit and vanish — it decays. LockIn90 models caffeine, alcohol and other inputs on real pharmacokinetic curves, so a 4pm coffee weighs on your 9pm sleep exactly as much as the chemistry says.
- Calibration confidence: 91% — model v4.2 · n=128 events.
- Days to calibrate: 30 — tuned to your physiology.
Caffeine concentration vs glycemic response
Caffeine decays on a pharmacokinetic half-life curve (t½ ≈ 5h) against a glycemic response curve, over 12 hours.
Patterns earn their place
A loop only becomes an insight once it clears a paired statistical test — so you act on evidence, not coincidence. α = 0.05 · paired t-test.
- Late caffeine → afternoon dip — 4 of last 5 days · n=22 · p = 0.012 · significant.
- Sub-7h sleep → next-day low — observed 9× · n=31 · p = 0.004 · significant.
- AM training → evening lift — observed 11× · n=28 · p = 0.031 · significant.
- High pollen → mood dip — observed 4× · n=19 · p = 0.21 · not yet shown.
Athletes stopped guessing.
“The attribution is the whole thing. It told me my 3pm crash was a 4pm-the-day-before espresso. Moved the cutoff, the dip vanished.”
“I coach twelve athletes. The privacy-gated view plus E2EE messaging means I see the trends without seeing their diary. That's the line I needed.”
“The 31-day forecast flagged a recovery crater before a travel week. I deloaded ahead of it instead of after. First time tech actually changed a decision.”
Know the why. Start free.
Connect a source, log a few moments, and get your first explained energy day — score, curve and the why — by tomorrow morning.
Start free