Features
See which interventions work for your body.
Outliyr goes after the root cause, so the fixes tend to pay off in places you were not even measuring instead of costing you somewhere else.
02The thesis
Root cause optimization produces off-target benefits. Biohacking has side effects.
Most platforms treat one symptom at a time. Each new pill creates a new problem. Outliyr investigates upstream so the same fix cascades into hormones, mood, glucose, recovery, and body composition.
The shape of thirty days.
One Outliyr member, running an upstream sleep, light, and breath protocol. The dashed shape is Day 1. The solid shape is Day 30. Each axis is one of their measured outcomes.
Statistics computed from the member's own daily Pulse, voice captures, wearable metrics, and bloodwork. Not a marketing testimonial. Not a clinical trial. A research dashboard reading the data.
03Inputs
Nine streams of biology. One workflow.
Outliyr ingests every signal you already track. Subjective, vocal, biometric, biological, behavioral, environmental. Each one is right about its slice and incomplete on its own. Joined, they tell the actual story.
04Mirror
Read the patterns your face and tongue already show.
Mirror is a guided phone-camera read, on every plan. Point the camera at your tongue and face, and Outliyr reads the observable patterns: coating, color, texture, the cues practitioners have looked at for centuries. It pairs two of the oldest observational traditions with what modern science can correlate, so an ancient read comes with a present-day reason to care.
05Dry fasting
Dry fasting, actually guided.
The fasting apps you can name are built for water fasts and run a timer. Outliyr guides the fast: researched stage content from the first hours through the day-4 window, matched to your protocol, with your biomarkers live on the lock screen the whole way.
Stage-by-stage guidance
Every stage explains what's happening and what to watch, hour by hour. Where the research is early, the card says promising instead of proven.
Biomarkers ride along
Connect a CGM and glucose flows in on its own. Ketones, GKI, weight, and quick logs sit beside the timer, on the lock screen via Live Activity.
Fasts become experiments
Link a fast to an N-of-1 experiment and get a graded verdict: Experimental, Under test, Confirmed. Your Personal Scientist reads your fast as context, so answers during hour 70 know you're at hour 70.
Dry fasting is an advanced protocol. In-app guidance stays conservative, adds advisory framing as fasts run long, and counts as education rather than medical advice.
06Outputs
Four surfaces that read your data back to you.
No more cross-referencing three apps and a spreadsheet. Outliyr says what your data is doing in plain English, with the math attached, refreshed every day.
07Trial engine
You tap three buttons. Outliyr does the heavy lifting.
No clinician-grade data entry. No setup wizard the size of a tax return. You mark when you start, you tap a button each day, you read the report when it closes. The engine handles the math, the baselining, and the side-effect watch behind the scenes.
- Auto-derives a baseline window from the equal-length pre-trial period of your data.
- Snapshots your supplement stack at the moment of start so mid-trial changes are flagged later.
- Locks the cohort of metrics being tracked so the report has a clean comparison.
- Side-effect watch on every connected metric, not just the ones you picked. Catches off-target signals you didn't predict.
- Auto-excludes life-event days you tagged (travel, sick, family emergency).
- Flags concurrent-trial confounds and any mid-trial stack changes that could muddy the read.
- Per-endpoint verdict: Improved / No change / Worsened / Inconclusive. Every metric you picked gets its own.
- Statistical confidence gating refuses to declare on too little data (n_days fewer than 10 stays Inconclusive).
- AI narrative synthesizes the verdicts, side effects, and adherence into a 2-3 sentence summary you can act on.
HRV. REM Sleep. Glucose. Each gets its own verdict.
Trials track unlimited primary endpoints in parallel. Pick the metrics that matter to your hypothesis. The engine runs each one through the same statistical pipeline (baseline mean, intervention mean, sigma delta, confidence gate) and reports one verdict per endpoint. The rest of the category reads your data and shows you patterns in it. Outliyr runs the trial, on any metric you log or sync, and grades each endpoint on its own.
08After the trial
Verdicts that stay verified.
Running the experiment is half the product. The other half is keeping the answer true: confirmation grades, post-trial monitoring, and a running read on what every bottle in your stack is earning.
And when there isn't enough data, Outliyr says so: zero-log drafts show Never Started, weak signals stay Experimental, and no verdict gets rounded up.
09Passive signals
Reads you never had to log for.
Some of the most useful signals come from data you already generate. Outliyr derives these in the background and only speaks up when something is worth your attention. No extra tracking, no wearable required for several of them.
Outliyr vs ChatGPT, Claude and Gemini.
Use them. Today's AI assistants can read connected health data and explain your labs well. The question is what you need on top. Outliyr exists for one job they don't do for you yet: running a fair test on your own body and keeping the result.
| What you need | ChatGPT, Claude, Gemini | Outliyr |
|---|---|---|
| Read your connected health data | Yes, through each one's health features (rollout, region and plan limits apply) | Yes: Oura, Whoop, Ultrahuman, Apple Health, Health Connect, labs, DNA |
| Explain labs and trends | Yes, and well | Yes, from your own baselines |
| Draft a protocol | Yes | Yes, and Experiment Autopilot builds it for you |
| Run the test for you: on and off phases, dose reminders, adherence, washout | Not described in their published consumer docs (October 2026); you run it yourself | Yes, that is the product |
| Keep a graded verdict per intervention, rechecked as data lands | Memory keeps context, not a verdict for each thing you test | Yes: Experimental, Under test, Confirmed, with Evidence Watch after the test |
| Bring your experiment results to your favorite AI | Only what you paste in or connect yourself | Yes. Data Passport exports work anywhere |
Why this holds up as AI gets better
Smarter models make explanations cheaper. They can't create measurements of you that were never recorded. The valuable thing is a clean record of what you changed, how consistently you did it, and what actually happened, measured the same way each time. Outliyr builds that record with you, and it stays yours: you can export it and take it to whichever AI you prefer.
When you don't need Outliyr
If your current AI setup already runs your experiments and keeps the results, keep using it. Outliyr is for people who want the test run for them, instead of maintaining their own tracker.
Sources, read October 2026: ChatGPT Health, Claude for health, Google Health Coach. Their features change often; check their current docs.
10Start
The Free tier is the product. Not a teaser.
Pulse, voice, wearables, bloodwork, the experiments engine, the Insights Engine, the BioHarmony Reports library, and the cross-platform apps on day one. Pro deepens cadence and history.
Outliyr runs as a web dashboard plus iPhone and Android apps on the App Store and Google Play, with your fast and your recovery on the lock screen.