Private beta · Austin

Your health data,
finally in context.

Cohort connects the apps you already use, matches you with eight people genuinely like you, and surfaces what actually moves the needle — from real variation inside your group, not population averages.

Join the waitlist See how it works

8 People per cohort
9 Connectable sources
0 Biomarkers we can read

The Cohort app home screen: a health score of 74 out of 100 with body composition, cardiovascular, performance and recovery sub-scores, above a cardiovascular card showing a 48 bpm resting heart rate, 68 ms HRV and a VO2 max of 52.

How it works

Three steps. No new habits to build.

01 2 minutes

Connect what you use

Apple Health, Strava, Chess.com, UDisc, blood panels, DEXA scans. Read-only, revocable, and you pick each one.

02 Monthly

Get matched into eight

Matched on age, baseline, and goal — not demographics. Cohorts re-sort seasonally as you change.

03 Ongoing

See what actually moves

Weekly insights drawn from natural variation inside your group, with the sample size always shown.

Your cohort

Compared to people who actually match you.

Percentiles against the general population tell you nothing useful. Cohort matches on age, baseline fitness, and goal — then shows you where you sit inside a group of eight, and how fast you're moving relative to them.

Mile pace, weekly volume, and consistency tracked against your cohort median — not a national average.
A leaderboard that stays small enough to mean something: eight people, ranked across connected domains.
Cohorts re-sort seasonally, so your comparison group keeps pace with your actual fitness.

The My Cohort screen: a twelve-week mile-pace chart plotting your improvement against the cohort median and the cohort's range, reading "your mile pace improved 8.2% since week 0 — 1.9 times the cohort median of 4.4%, top 2 of 8", above a Where you stand table comparing your pace and weekly runs to the median.

Insights

Patterns, not prescriptions.

We surface what we observe in natural variation across your cohort — and we always show the sample size behind it. Every finding is tagged for what it is. A pattern in eight people is a lead, not a law, and we say so.

Every finding carries its sample size and window — n = 6 of 8, 9 weeks — so you can weigh it yourself.
Cross-domain patterns connect what nobody thinks to compare: sleep against chess rating, recovery against pace.
Insights graduate as your history deepens — personal trends first, cohort findings once there's enough signal.

The Insights screen: a cross-domain finding that cohort members who added a weekly Zone-2 run improved mile pace 1.9 times faster — plus 7.9% against plus 4.2% — tagged "n = 6 of 8", "9 weeks" and "pattern, not proof", with the caption that this is a naturally occurring split rather than a prescription.

Privacy

"We cannot see your blood pressure" is a technical fact, not a policy promise.

Encrypted biomarker vault

Blood panels and body scans are encrypted on your device. We store ciphertext only — the keys never leave your phone.

Two-tier consent

Personal insights and cohort contribution are separate switches. Turn either off without losing the other.

Anonymized contribution

What reaches your cohort is signal, never records. No names, no raw values, no reverse lookup.

Cohorts form monthly. Get in the next one.

Private beta in Austin. Connect the apps you already use — setup takes about two minutes.

WAITLIST OPENS SOON · WE'RE NOT COLLECTING EMAILS YET

Your health data,finally in context.