Some things that came to mind that you could fix quickly if you want to: You gotta publish bullet-proof privacy policy and terms of use. You also need to disclose what underlying API you're using (Anthropic, OpenAI, something else?). I'd personally stay with US API providers because of their strong "no using your data" policies. However, that could change at any moment, so you have to be careful.
Positives: Looks great! The use case particularly resonates with me, but that's not everyone's cup of tea.
Suggestion: Instead of me having to ask questions, the app should surface divergent metrics on its own. That's the unlock imho!
Great work, keep it up!
At onboarding, we do provide a term sheet that a user can download.
Regarding API, we use varying models depending on the use case, but yes they're either Openai or Claude. Since everything is hosted on Microsoft Azure, everything is routed through their Foundry.
Regarding asking questions, it does surface insights for wearables daily. For lab records, that's what is being added right now.
Thank you so much for your feedback! Please do register if you'd like to try it out. I would really appreciate feedback on the app itself, whether you use it individually or with others too.