We want to make getting insights less effort and more practical.
You guys should really focus on AI implementation to boost the slow improvement of the app.
I’d love to see a list of all correlations, including by time, sorted by strength. For example, X has a 90% correlation with Y at 3 days offset. Then I’d like to be able to ignore certain entries, because some correlations aren’t valuable to me. This would allow me to automatically see new interesting correlations as they were discovered.
The “daily trends report” is almost completely useless, ands makes me depressed rather than being helpful, because i see the same exact items every single day. I’d like to be able to check them off or something so the same things don’t reappear the next day. It would be great if the suggestions were more detailed- what is correlated with a particular symptom that has been a challenge lately, rather than what is correlated with the entire symptom score, esp if that is going to be the exact same things every single day
I’ve seen AI mentioned in another comment. I wouldn’t feel comfortable with an AI implementation just because every other app does it as well. It could be very useful for generating stats and correlation, but assuming you would go with a backend like Claude, this could cost a lot of money. And so have a direct impact on the subscription pricing.
The way I see things, if someone wants to process their data through any kind of AI/LLM, the text export is more than enough to ask Claude or else to analyse said data. All of this independently from Bearable. And I’m not even talking about GDPR stuff here. AI companies likes OpenAI don’t have a good reputation when it comes to data privacy. And implementing a custom LLM backend would probably be too much work for the Bearable team.
Long story short: unless it is useful and affordable, AI integration isn’t necessary.
Yes! That would be a major help being able to take any of the factor lists and marking it with “hide” – a lot of them are redundant or not-applicable in certain views, so I’d go as far as making the “hide” lists align with the current report being viewed, the factor list filters applied or any other user-filterable view.
Example use case: - Impact View - How: Factors Impact: Sleep Quality on: Same Night
All “hidden” items could be revealed either in a special section below (accordion that starts collapsed by default) or with a “unhide” toggle (which has the benefit of returning it to it’s position in the list).
The “Hide This” actions could be “Hide With This Filter / From This Report” (demonstrated in example) or “Hide From All” (for things you want to track but don’t want to see).