Your Garmin gives you a recovery score every morning. Do you actually know what to do with it?
Mine says 68 today. Green zone apparently. But 68 out of what? Compared to what baseline? Why 68 and not 71? Should I train hard or take it easy? The app doesn't say. It just shows me the number and a vague color.
Most wearables give you a score with zero explanation of how they got there. Black box. Proprietary algorithm, tuned for some average user that probably isn't you. You either trust it or you don't, but you have no way to verify it either way.
And if you're building a health product, this gets worse. Your users come with 6 different devices. Whoop gives a recovery score, Garmin gives a Body Battery, Oura gives a Readiness Score. Three different algorithms, three different scales, no way to compare them. You either pick one device and exclude everyone else, or you build your own normalization layer from scratch.
We launched Open Wearables on Product Hunt today to fix exactly this. It's an open-source platform with one unified API for 10+ wearable providers and open health scoring algorithms that work the same across all devices. Every formula is public, auditable, and customizable. If your users are athletes, tune the thresholds for athletes. If you're building for clinical use, your medical team can verify every calculation.
The scores are just the start. We also built a Health AI Engine that actually reasons about the data: it detects trends, flags anomalies, connects patterns across scores, and produces recommendations a coach would give. Not "your HRV was 45" but "strain exceeded capacity 3 of last 5 days, reduce intensity, estimated recovery to baseline: 2 days."
MIT licensed. Self-hosted. $0 per user.
Would love your support and any feedback on what you'd want from something like this.
Replies
TinyCommand
This seems interesting. Does it also take into account medical conditions like blood pressure, diabetes, insulin resistance, etc.?
@priyanka_gosai1 I’ve honestly been confused by those scores too. I usually ignore them, but your approach makes me want to rethink that.
@priyanka_gosai1 @deangelo_hinkle Always felt those recovery scores were a bit arbitrary. I see the number, but I don’t really change my behavior based on it.
Open Wearables
@priyanka_gosai1 @deangelo_hinkle @henry_lindsey If you can't trace why the number is what it is, acting on it feels random. A score that doesn't explain itself doesn't change behavior, it just adds noise.
@priyanka_gosai1 @deangelo_hinkle For me, the biggest issue has always been trust. I get a number, but I don’t know what inputs drove it or how sensitive it is to changes.
Open Wearables
@priyanka_gosai1 @deangelo_hinkle @lakeesha_weatherwax Input visibility and sensitivity are exactly what the open algorithm gives you. You can see which inputs carry the most weight and simulate how a change affects the output. That's what turns a number into something you can act on.
@priyanka_gosai1 @deangelo_hinkle If I sleep slightly better, does that move the needle meaningfully? I like that you’re making the logic visible and adjustable.
Open Wearables
@priyanka_gosai1 @deangelo_hinkle @shawn_idrees With an open formula, you can check that directly. Sleep duration, consistency, and stage distribution each carry different weights, and you can trace how much a 30-minute improvement in deep sleep actually shifts the score. That's what makes it worth paying attention to.
Open Wearables
@priyanka_gosai1 Not yet. OW focuses on fitness wearable data: activity, sleep, HRV, recovery, and strain. Medical condition data like glucose or blood pressure from clinical devices isn't in scope today, though the architecture is extensible and providers like CGMs could be added down the line.
Building in cycling analytics. The reason cycling figured this out is the Coggan framework. FTP, W/kg, TSS, CTL all map to specific prescriptions that are an industry standard, so the number isn't just a number, it tells you whether to do intervals today or sit on the couch. That only worked because the formulas are public and reproducible across platforms. Curious to see whether Open Wearables ends up taking recovery scores in the same direction, where an open algorithm carries an open recommendation, not just an open number.
This is a massive pain point for anyone building in the health-tech space. Highly appreciate you guys opening up the black box.
Dealing with the disparate "editorial choices" and proprietary metrics between Oura, Garmin, and Apple HealthKit can be an absolute nightmare when trying to architect a cohesive user experience.
For the platform I’m currently building, I actually took a slightly different architectural route to bypass this noise: we use medical blood tests as the rock-solid clinical benchmark, and treat wearable metrics strictly as contextual, behavior-driven trends rather than absolute truth. Because at the end of the day, a perfect morning readiness score doesn't mean much if your metabolic labs tell a different story.
That said, having a unified, open-source API layer like Open Wearables to standardize the raw inputs without the proprietary smoothing is a game-changer for infrastructure. Definitely going to look into integrating this to streamline our data ingestion pipeline. Congrats on the launch!