Hey all, I'm currently a solo iOS developer who just released my first app, Timbre, onto the App Store (iPhone only for now). Here's a demo to show you how it works:
https://www.youtube.com/watch?v=ALoPAFb31Bk.
Timbre originally came about because I wanted to build a sleep sounds & sleep tracker app. I had used Headspace, Calm, Pillow, etc. and figured that instead of paying $70 / year I could just build my own app and pay $100 / year (not a great plan in retrospect). As I started building Timbre, I realized that one of the biggest issues I had with these other apps is that I could never really see which sounds genuinely helped me sleep. With that in mind, I pivoted on my original idea and decided to really hone in on connecting sleep quality and sounds.
Using a custom ML model, Apple's CoreML framework and HealthKit's sleep stage data, I made it so that Timbre could calculate a personal sleep quality score. Therefore, when you record a sleep session while listening to a sound, the app can rank the sounds you've listened to based on your sleep quality scores over time.
Of course, the app also has a smart alarm, offline support, sleep analysis and more, but I'd love to get some feedback & suggestions from the demo video. Also feel free to ask me any questions and I'd be happy to answer them!
Can you elaborate what data are you using for ML purposes? Specifically, what permissions is the iphone requestion to collect data ?
What exactly is allowing you to make assertions about quality to sleep, except logical stuff like ambient sounds like snoring, etc?
Do i need an apple watch??