We built this because we needed it.

I'm a triathlete and a growth engineer. I came up through endurance the way a lot of people do: YouTube got me to my first triathlon, a real coach got me to my second. Then the bill arrived. I switched to static plans off the shelf - cheap, structured, and exactly as useful as your week is predictable. The moment life got in the way (a work trip, a bad couple of nights of sleep, a stretch of long days), the plan on the calendar was already wrong, and I was the one left guessing.

A static plan is a snapshot of what should happen if everything goes right. Real training is never that. You nail a workout and the next day says “easy” when you could push. Your HRV craters and it says “intervals.” A good human coach reads the situation and adjusts. A PDF can't. And paying a coach 200-400 dollars a month to send adjustments over text isn't realistic for most of us.

athletedata.health is what we wished existed. It sees what actually happened - the session, the sleep, the strain, the meal - and adjusts what's coming next. It catches the drift before the plan and the athlete diverge. For multi-sport especially, where swim, bike, run, gym and recovery all stack into one fragile schedule, having one thing that holds the whole picture is the difference between making it to the start line and not.

If you've ever been mid-block and realized the plan no longer matched the athlete on the page, we built this for you. If there's a sport, a metric or an integration we don't handle yet, message me. I read every one.

I'm an AI engineer and endurance athlete in San Francisco, training for a marathon. Most of my time goes to running and lifting. I track everything: Strava for runs and rides, Hevy for gym sessions, WHOOP for sleep and recovery, Withings for body measurements.

The problem was always context. I'd screenshot my Strava summary or copy my WHOOP recovery score and paste it into Claude or ChatGPT to get advice. Every conversation started from zero: no history, no profile, no awareness of what I did yesterday or how I slept last night. It worked, but barely.

So I built athletedata.health. It connects directly to all of those platforms and gives the AI real-time access to my actual training data. When I finish a run, it messages me on Telegram with analysis. When I wake up, it checks my sleep and recovery and tells me if I should adjust the plan. I can snap a photo of my meal and it tracks the macros. The coach remembers everything: my goals, my injury history, my preferences, what we talked about last week.

I'm sharing it because other people have the same problem. If you use any combination of fitness apps and want an AI coach that actually knows your data, this is it. And if there's a platform you want connected or a feature you'd find useful, reach out. I'm happy to build it.