Private body-composition intelligence
One weekly verdict from the scans you already trust: what moved, whether it is real, and the next right decision.
Reviewed scans, your goal and the context you log go in. One verdict comes out, with the reason beside it and what to do before the next scan.
Weekly readMay 14 – May 21, 2026Sample week, fictional data
Mar 12 · 16.6%Body fat, five scans, inside its noise bandMay 21 · 15.5%
Five questions, answered every week.
What changed?
Fat mass ↓ 0.2 kg · Body fat ↓ 0.2 % · Lean mass ↓ 0.2 kg
Is it signal or noise?
Noise. Every change this week sits inside its expected scan-to-scan band.
What likely caused it?
The deficit is selective: fat mass is moving down while lean mass holds within InBody's precision band.
Are we trending toward the goal?
Body fat 16.6% to 15.5% across five scans, Mar 12 to May 21.
What is the next best decision?
Fat mass down again, lean held, keep the cut signal running.
- Body weight85.7kg↓ 0.4 kgNoise range
- Body fat15.5%↓ 0.2 %Noise range
- Fat mass13.3kg↓ 0.2 kgNoise range
- Lean mass72.4kg↓ 0.2 kgNoise range
The verdict for this week
Hold
Hold the protocol. Rescan in 7 days.
How a scan becomes a decision
Three checks come between a scan and its verdict.
OWLiFT reviews the values, matches the scan to the last one and sets the change against the noise. Follow one sample scan through, then change what happened to it.
Scan review
Confirmed
Were the values read correctly?
You confirmed the values from your May 21 InBody scan. Only confirmed or corrected scans feed a verdict.
Scan review
Needs review
Were the values read correctly?
The scan is saved and waiting for you. Confirm or correct it before it can feed a verdict.
Scan review
Confirmed
Were the values read correctly?
You confirmed the values. Your first confirmed scan starts the record everything else is read against.
Comparability
Matched
Can it be set beside the last scan?
Same device as your May 14 scan, and this one was taken fasted, in the morning. The same device under known conditions is what makes two scans comparable.
Comparability
Left out
Can it be set beside the last scan?
The waiting scan is not compared with anything. The read keeps to the last scans you confirmed.
Comparability
Nothing to match
Can it be set beside the last scan?
The first confirmed scan, so there is nothing to compare it with yet.
Interpretation
Noise range
Is the change bigger than the noise?
Fat mass is down 0.2 kg and body fat 0.2 points, and lean mass moved 0.2 kg. Every change this week is inside the expected scan-to-scan noise, so none of them is a signal on its own.
Interpretation
Not read
Is the change bigger than the noise?
Nothing in the waiting scan is read. The read you already have stays as it was, built from your confirmed scans.
Interpretation
Not enough data
Is the change bigger than the noise?
One scan cannot show movement. A trajectory appears once there are two confirmed scans to read movement from.
Decision
What happens next?
Verdict: Hold
Fat mass down again, lean held, keep the cut signal running.
Hold the protocol. Rescan in 7 days.
Confidence: Low. 5 trusted scans inform this read. Main caveat: InBody fat-mass readings carry roughly ±0.5 kg noise; a single week's change sits inside that band.
Decision
What happens next?
The last read stands.
The scan waits for you. Confirm or correct the values, and it counts toward your next read.
Decision
What happens next?
No verdict yet.
Log a second confirmed scan. Two scans let Hoot read what changed, not just a snapshot.
The matched case is the sample week, on fictional data. The other two show what the product does.
The launch film
What counts.
The trend
A change has to clear the band.
Every confirmed scan sits on one trend, drawn against the range the measurement normally wobbles by. A move inside the band is noise. A move outside it is worth another look.
Body fat, five scansBand ±0.5 points
May 21 scan, 15.5%. InBody · fasted · morning Same device as the previous scan.
- Confirmed scans only
- The trend is drawn from scans you reviewed, and from nothing else.
- The band is drawn
- Expected variation is drawn on the chart, so a wobble is easy to tell from progress.
- The newest scan is named
- The latest reading says which device it came from, how it was taken, and whether that device is the same as for the scan before.
Context
Context is what tells a real change from noise.
Sleep, supplements, protocols and symptoms are read alongside the scans. None of it is required. What is missing caps confidence in the weekly read. It never blocks it.
- Journal
- Sleep, stress, training and mood, day by day.
- Stack
- Vitamins, minerals and supplements, kept as dated records.
- Protocol
- Medication and other protocols, with the date each one changed.
- Symptoms
- What you notice, kept as context. Never diagnosed.
The four verdictsThis sample week: Hold
- Hold
- Adjust
- Wait
- Investigate
The decision
One decision, measured against your goal.
Set the target once. Every weekly read is judged against it, and the answer is always one of four words: Hold, Adjust, Wait or Investigate, with the reason beside it.
The four verdictsThis sample week: Hold
- Hold
- Adjust
- Wait
- Investigate
- Each measurement on its own
- Every row has its value, its change and one plain class for it, such as Stable, Possible signal or Noise range.
- The reason beside it
- Hoot says what it reads in each row and why it matters.
- One next step
- Each read ends with what to do before the next scan.
Ask Hoot
Hoot shows how a verdict was reached.
Ask why, and Hoot lays out what it looked at, what was missing and what it refused to assume. The panel is built from the saved read in front of it.
Ask Hoot in the demoHootSample week, fictional data
Why did Hoot call this week the way it did?
What it looked at
- Latest scan logged at high source confidence.
- Both scans taken on the same device.
- 5 trusted scans inform the trend.
What limits it
- InBody fat-mass readings carry roughly ±0.5 kg noise; a single week's change sits inside that band.
- Appetite suppression is self-reported, no food log to corroborate intake.
- Only one week at the 5 mg dose; the trend is early.
Why
- The deficit is selective: fat mass is moving down while lean mass holds within InBody's precision band.
Privacy and limits
Private by default. Plain about its limits.
The short version of the privacy policy, the terms and the disclaimer. The full text is one click away.
Your scans are yours.
Scans, bloodwork and context are stored against your account. They are not shared with other users, and each account can see only its own records.
Privacy policy
Nothing is sold. Nothing trains outside models.
Your data is not sold, and it is not used to train external AI models. A small number of hosting, database and AI providers process it on our behalf to run the service.
Privacy policy
You confirm every scan before it counts.
Values read from a report wait for your review. Nothing is saved until you confirm it, and unconfirmed values never reach a verdict.
Log Scan and the weekly read
Take it all with you, or erase it.
Export your scans, journal, goals, weekly verdicts and Hoot conversations from Settings. Delete your app data or your whole account, each behind a typed confirmation.
Settings
It is not a doctor, and it says so.
OWLiFT interprets reviewed scans to support your own decisions. It does not diagnose, treat or prescribe, and it does not replace a clinician. Scans are noisy and Hoot can be wrong, so check both against your own data.
Disclaimer
It is a beta, and it claims no more.
OWLiFT is pre-release and invitation-only. It does not claim a privacy or security certification, or clinical-grade accuracy.
Terms, privacy policy and disclaimer
The next step
Make the next right decision.
See it on sample data first. Free during the beta, no card required.
Your own account is by invitation while OWLiFT is in beta.