Saksham

researcher @CarnegieMellon | indie maker e/acc

[Saksham @vu3dtu]

I loved the idea behind the Google Fitbit Air: an LLM wrapped around your health data, daily briefs, and a coach you can ask questions.

But there app is really terrible, it's expensive $100 band plus $10/mo, and Google getting a constant stream of your heart rate, sleep, and other private data. Whoop is worse, with a subscription that runs up to $360 a year.

So I bought a $7 generic Chinese smart ring from Temu, reverse engineered its BLE protocol, and built an app around it.

Introducing PulseLoop: no subscription, open-source iOS app. Your health data stays on your phone, paired with an AI coach that reads your real ring data, draws charts, and remembers context. Free, bring your own API keys. Demo and code below.

[Lewis Hamilton @LewisHamilton]

REMEMBER WHO YOU ARE

[Andrej Karpathy @karpathy]

In awe of SpaceX and its story - past, present and the future. You can think about it in 10+ different ways and continue re-blowing your mind in circles. Huge congrats to the team! 🚀

[Saksham @vu3dtu]

@arkhelion I haven't thought much about it yet - but it might be good to get confidence in the ring values (if it's close to the apple health data). If we have high confidence keep the data point, otherwise get rid of it. Do you have some ideas?

[Saksham @vu3dtu]

I loved the idea behind the Google Fitbit Air: an LLM wrapped around your health data, daily briefs, and a coach you can ask questions.

But there app is really terrible, it's expensive $100 band plus $10/mo, and Google getting a constant stream of your heart rate, sleep, and other private data. Whoop is worse, with a subscription that runs up to $360 a year.

So I bought a $7 generic Chinese smart ring from Temu, reverse engineered its BLE protocol, and built an app around it.

Introducing PulseLoop: no subscription, open-source iOS app. Your health data stays on your phone, paired with an AI coach that reads your real ring data, draws charts, and remembers context. Free, bring your own API keys. Demo and code below.

[Saksham @vu3dtu]

@signulll @netcapgirl start by reply to my dm

[Saksham @vu3dtu]

True to some extent. It's also harder to do step counting from rings as a lot of fine finger movement can be confused with steps. HR estimation is using PPG sensors that changes in absorption of light as result of blood flow, quality of the sensor and the algorithm to detect peaks in the signal can affect the estimates. I have a lot of academic work on PPG :)

[Saksham @vu3dtu]

I have been wearing apple watch, fitbit air and the ring for last 2 weeks. There is not a lot of variance in the steps, heart rate (±5 bpm) and sleep (±10 minutes) surprisingly. Although a lot of second order derived metrics are harder to estimate like calories burnt. I have also been testing slightly more expensive rings $20-30 and it is even closer. I am also working on integrating with apple health, thay should also help the accuracy as it can get some data from your phone!

[Saksham @vu3dtu]

@0xNeoArch Yes, I would love all the help. Not sure about the version. The sellers don’t really have any standardised naming patterns. But adding support shouldn’t be hard, protocol should be pretty much the same.

[Saksham @vu3dtu]

@Timothy98537991 Go for it! Feel free to make PRs with android port.

[Saksham @vu3dtu]

@TedRozycki I am working on it, it’s not as easy to reverse engineer the data protocol of fitbit.

[Saksham @vu3dtu]

@SorinAmzu https://t.co/5J15NJzuvo

[Saksham @vu3dtu]

I have been wearing apple watch, fitbit air and the ring for last 2 weeks. There is not a lot of variance in the steps, heart rate (±5 bpm) and sleep (±10 minutes) surprisingly. Although a lot of second order derived metrics are harder to estimate like calories burnt. I have also been testing slightly more expensive rings $20-30 and it is even closer. And, yes I am working on apple health integration!

[Saksham @vu3dtu]

I have been wearing apple watch, fitbit air and the ring for last 2 weeks. There is not a lot of variance in the step, heart rate and sleep surprisingly. Although a lot of second order derived metrics are harder to estimate like calories burnt. I have also been testing slightly more expensive rings $20-30 and they are even better.

[Saksham @vu3dtu]

@netrunner_btc I agree, they have some good derived metrics. If there were to be a big open source project we can write algos for these derived metrics. There is a lot of academic research around this.

[Saksham @vu3dtu]

Yeah for sure, they have a lot of derived metrics and there measurements will also be more accurate (better sensors, more data). But I like to think of this as "pièce de résistance", you have absolute control over your data, and you pay for what you use albeit at the cost of accuracy.

[Saksham @vu3dtu]

Google probably has 200 engineers working on that, and the main page of the app is just LLM LLM LLM LLM. It makes a lot of mistakes and for me it stops measuring some vitals and steps from time to time.

I made this app alone! There is a lot of room for improvement, help me make it better! But fundamentally you keep your own data and don't pay for any subscriptions.

[Saksham @vu3dtu]

@synopsi makes sense!

[Saksham @vu3dtu]

This is the new version hand of god!

[NEW YORK KNICKS @nyknicks]

THE MOMENT.

[Saksham @vu3dtu]

@aayush_mantry I am working on adding data import/export from apple health!

[Saksham @vu3dtu]

@theknightqueen I am working on trying to reverse engineer Fitbit data, will get back to you soon!

[Saksham @vu3dtu]

@chunky_bear_ @blinkenlicht It's not just about the money, they own all your personal health data! Just wait until they start selling it to insurance companies!