An onboarding chat with an AI Health Coach powered by Gemini collects goals, routines, and obstacles to generate a personalized wellness plan in about five minutes. The plan includes suggested workouts and targets that can be manually adjusted or refined through follow-up chats. The experience feels approachable rather than overly clinical, with the coach driving daily engagement more than the tracker alone. Morning check-ins provide sleep recaps, post-workout summaries appear after exercises, and nightly overviews connect activity, recovery, and stress. Messages often end with questions that prompt natural chat instead of dismissible notifications. Activity detection is generally reliable, with occasional misreads that can self-correct using heart-rate data, and improved recognition after manual logging and feedback.
"Setup begins with an onboarding chat with the new AI Health Coach, powered by Gemini. It asks about your goals, routines, and obstacles before generating a personalized wellness plan. Depending on how much detail you share, including the option to upload medical records, the process takes around five minutes. From there, the app generates a weekly plan with suggested workouts and targets that you can tweak manually or refine through follow-up chats with the Coach."
"It sends you check-ins in the morning with sleep recaps, post-workout summaries after exercises, and nightly overviews that connect your activity, recovery, and stress levels into something more coherent. Most of these messages also end with a question about how you're feeling, which naturally opens into a chat rather than feeling like another notification to dismiss."
"Automatic activity detection is solid overall. The Air consistently recognized walks and even generated useful summaries about intensity and recovery afterward. I haven't run into any workout hallucinations (yet), though there were occasional misreads. On one day, for example, the Air logged a walk as a run but then immediately followed with a note pointing out that my heart rate data suggested it was probably a walk."
"The detection algorithms also noticeably improved with feedback. During my first three days of testing, the Air missed a recurring high-intensity workout class. But after I manually logged the sessions a few times, it began recognizing them automatically. Like the Oura Ring, the Air gets smarter the more context you give it."
#ai-health-coaching #personalized-wellness-plans #activity-tracking #workout-detection #user-feedback
Read at www.wired.com
Unable to calculate read time
Collection
[
|
...
]