What Can AI Do With Your Strava Runs and Rides?

Strava has no direct API link into Wellness Project, and that is by design rather than a gap. Strava already syncs your runs and rides into Apple Health on iOS or Google Health Connect on Android through its own native integration, the same pathway Strava has offered for years to feed the phone-level health record on both platforms. Wellness Project pulls your activity history from that unified layer instead of talking to Strava directly, which means the connection is a one-time setup, not a tutorial you repeat every time you want an answer.

The practical effect is that once Strava is flowing into Apple Health or Health Connect, every run and ride shows up in Wellness Project next to your sleep, heart rate, and nutrition logs, all in one history that Claude and ChatGPT can query. This same relay path is how Garmin, Whoop, Withings, Peloton, Coros, Wahoo, Amazfit, Polar, and Samsung data reaches Wellness Project too, since none of those devices have a direct link either. For the exact setup steps, from opening Strava's connected apps settings to confirming your first synced activity appears, see the Strava connect guide, and for what the same relay layer can do beyond just Strava, see the guide to what AI can do with Apple Health data.

Once activities are syncing, the questions worth asking are the ones a spreadsheet of individual runs cannot answer on its own. A few that work well:

• "How has my average pace changed over the last 8 weeks?"
• "Did my heart rate drift upward during my long run on Saturday?"
• "What was my total mileage this week compared to last month?"
• "How much elevation gain did I log across this training block?"
• "Did my sleep quality drop the week I ran my highest mileage?"
• "Compare my average heart rate and duration on my last two long rides."
• "How many calories did I burn running this month versus last month?"
• "Am I trending toward overtraining based on my weekly distance and resting heart rate?"

These stick to the data types that actually make it through the Apple Health and Health Connect relay: distance, pace, duration, heart rate, elevation, and calories. Strava-specific fields like power or cadence are not guaranteed to pass through that relay, so questions grounded in distance, time, heart rate, and elevation get the most reliable answers.

A real answer names specific numbers and dates pulled from your logged activities, not a canned training tip. Ask about your pace trend and you get something like an average pace for each of the last several weeks, called out against the weeks where it improved or slipped, rather than a general note that consistency improves speed. Ask about a single run and you get that run's actual duration, distance, and heart rate figures compared against your recent baseline.

The bigger difference shows up on cross-domain questions. Because Wellness Project also holds your sleep, HRV, and nutrition logs in the same history, the AI can connect a mileage jump to a recovery dip or a missed protein target in a single answer, something Strava's own activity feed has no way to do since it only sees the run itself. A question like whether a 31-mile week left you under-recovered gets answered by looking at what actually happened to your resting heart rate and sleep duration that week, not by a generic rule of thumb about training load.

Strava is genuinely good at what it was built for: per-activity splits, segment leaderboards, kudos, and the social side of a single run or ride. None of that goes away when you connect Strava to Wellness Project, and this guide isn't arguing you should stop opening the Strava app after a workout.

Where a Strava AI coach earns its keep is everything Strava's activity-by-activity view wasn't designed to show: longitudinal trends across weeks or months, and reasoning that spans domains Strava doesn't track at all, like sleep or nutrition. Wellness Project's running specialist, Coach Elias Kiptoo, is one of eight named AI coaches available once your data is flowing in, and he can weigh in specifically on pace and training load questions with the same run history in front of him that you'd see in Strava, just read alongside everything else you've logged.