Atlas

My wife wears an Oura ring. She also constantly complains that her data is off. Sleep scores that don't match how she felt. Recovery numbers that contradict everything else her body is telling her. The ring became less a tool and more a thing she sometimes argued with.

Meanwhile, I have an Apple Watch Ultra. It's the most capable health sensor I've ever owned. HRV. Sleep stages. Workout intensity. Heart rate at a thousand different moments throughout the day. All of it sitting in Apple Health.

And it tells me almost nothing.

That was the problem I kept hitting. Apple Health is a database, not an analyst. It will show you your steps. It will show you your sleep duration. It will not tell you whether the way your HRV interacted with your sleep debt this week means you should push hard today or pull back. It will not tell you that your training load has crept past the threshold where injury risk starts climbing. The data is there. The interpretation isn't.

So I built Atlas.

Atlas syncs with your Apple Watch HealthKit data on a server, runs statistical analysis using current sports science research, and gives you back recovery, stress, and readiness scores. Modern training analytics with every input on the table, so you can see exactly what's driving every number.

Three things I cared about while building it.

First, transparency. Every score has its inputs visible. If your readiness drops 20 points, you can see exactly which components moved and by how much. No black box. No mystery weighted blend. The whole point of using a tool like this is to learn how your own body works, and a number you can't decompose is just another thing to argue with.

Second, modern methodology. The analytics aren't pulled from a 1990s VO2max table. They're grounded in current sports science research: state-based HRV analysis (sleep HRV and wake HRV are different signals, most apps blend them), ACWR for training load, multi-component recovery modeling. The kind of stuff coaches and labs use, not the consumer-grade simplifications.

Third, privacy. I'm not selling your data. I'm not training an LLM on it. Your HealthKit data goes to the Atlas server so I can run the analytics on years of history, but it's never shared with third parties or used to train external models. The only outside calls are Google and Apple OAuth for sign-in. Health data is the most personal thing you have, and Atlas is built so it stays that way: yours, used only to give you analytics.

Atlas is live on the App Store:

https://apps.apple.com/app/atlas-body/id6760951718

More about the project at:

https://atlasbody.co/

If you have an Apple Watch and you've ever looked at your Apple Health data and wondered what it actually means, I'd love for you to try it and tell me what you think.

What I'd most appreciate feedback on:

• Does the privacy framing feel clear, or do you want more detail on what's stored server-side and what isn't?

• Is the "every input visible" angle landing, or does the average user just want a single number?

• Anything in the onboarding that confused you?

I'll be in the comments.