For biohackers
The biohacker app that runs real N=1 trials.
Max designs the trial: one intervention, one outcome, a fixed duration. The data layer captures HRV, sleep, supplements, and labs. At the end of the window, you see the before-vs-after in your numbers, not in your memory.
Free · Now on iPhone · Syncs Apple Health, Fitbit & Oura
The result
Baseline, intervention, endpoint. Then a verdict.
Once the trial window closes, the result view stacks your baseline week against the intervention weeks side-by-side. Max calls signal vs noise.
Pick one supplement
The intervention is a specific supplement, magnesium glycinate, creatine, ashwagandha, whatever you actually want to test. Adherence is read straight off your daily supplement log.
Pick your baseline window
A fixed 14-day baseline runs against the intervention. Use your previous 14 days of already-logged data (intervention starts today) or collect 14 days forward (intervention starts in two weeks).
Set the intervention length
Anywhere from 14 to 120 days. Long enough for biology to react, short enough that you don't lose interest. The engine tracks adherence the whole time.
Every metric tracks automatically
Sleep, HRV, RHR, weight, body composition, daily steps, energy, focus, mood, whatever you log day-to-day rolls into the receipt. No metric-picker, no missed signal.
Read the receipt
When the window closes, the receipt shows the delta on every tracked metric. Below 80% adherence the receipt won't render, Max won't read an intervention you barely did.
What's inside
Quantified-self tooling that actually fits together.
HRV in context
Daily HRV against personal baseline, training load, alcohol, and sleep. Max reads it as one composite signal, not isolated readings.
CGM-aware in conversation
Wellness Project does not yet have a dedicated CGM module. You can discuss your time-in-range, post-meal AUC, and variability with Max from your CGM app screenshots and notes.
Stack tracker with start dates
Pin start dates on every compound. Weeks later the chart against HRV / RHR / sleep / lifts tells the story (or it doesn't).
Lab-anchored protocols
Atlas-designed protocols have explicit success metrics. ApoB target, HRV baseline shift, HbA1c trajectory, all defined upfront.
MCP for Claude, Custom GPT for ChatGPT
Talk to your data from inside Claude (native MCP) or ChatGPT (Custom GPT). Wellness Project is the structured backend; your AI is the universal interface.
Honest about noise
Max is structurally skeptical. Trials require a real baseline. He'll tell you when something is noise, not signal.
From the blog
Related posts from the crew
Max Kline · Aug 2, 2026
How Many Hours Before Bed Should You Stop Drinking Coffee?
The widely quoted 8.8-hour coffee cutoff comes from a 2023 meta-analysis, but a 2025 polysomnography crossover trial found 100 mg of caffe…
Max Kline · Jul 26, 2026
Does Drinking Earlier in the Evening Actually Protect Your Overnight Heart Rate?
A 20,968-person wearable study found each extra drink tracked with about 2.8 bpm higher overnight resting heart rate and 3.8 ms lower HRV,…
Atlas Mercer · Jul 23, 2026
Can You Trust Your Oura or Whoop Readiness Score?
When five wearables were tested against a clinical ECG across 536 nights, the raw nocturnal heart rate and HRV signals proved accurate to …
FAQ
FAQ
How does it compare to Heads Up Health, Stronger U, etc.?
Is the N=1 module live today?
Can I export everything?
Does it work with my CGM?
Stop chasing noise. Start running trials.
Free during early access, no credit card, iPhone, Android, web
Wellness Project is informational software, not a medical product. AI advisors do not diagnose, treat, or prescribe; for anything clinical, consult a qualified clinician.
