Oura + CGM experiment
A biohacker posted a 'vibe coding' experiment that combines Oura ring metrics with continuous glucose monitor data to calculate macros, insulin needs and glycemic control — a live example of wearables feeding personalised models. The thread highlights practical API use cases for symptom tracking and shows how consumer sensors can be stitched into ML workflows (x.com) (x.com).
A smart ring and a glucose patch measure two different parts of the same day. One watches sleep, stress, temperature, and heart signals from your finger, while the other sends sugar readings from fluid under your skin to your phone in near real time. (ouraring.com) (fda.gov) The glucose side is called continuous glucose monitoring, which means the sensor keeps checking instead of waiting for a finger-prick snapshot. The United States Food and Drug Administration says one approved system sends a reading every five minutes and can be used to help decide when to take insulin or carbohydrates. (fda.gov) The ring side is less about sugar and more about recovery. Oura says its Readiness Score runs from 0 to 100 and blends sleep, recent activity, resting heart rate, heart rate variability, and body temperature into one daily signal. (ouraring.com) Heart rate variability is the tiny change in timing between one heartbeat and the next. Oura measures it in five-minute samples during sleep, and the company says higher or lower values only make sense against your own baseline, not somebody else’s. (support.ouraring.com) Once those two streams exist as data, the experiment is mostly plumbing. Oura’s cloud application programming interface offers daily summaries and time-series data through OAuth 2.0 login, which is the standard permission system apps use to pull data from another service without sharing your password. (cloud.ouraring.com) That is what made the post interesting: it treated consumer wearables like ingredients, not finished products. Instead of stopping at the app dashboard, the builder stitched ring metrics and continuous glucose monitor readings into a custom model that tried to estimate macros, insulin needs, and glycemic control from one personal data trail. (x.com 1) (x.com 2) Oura and Dexcom have been moving in the same direction at the product level for months. On November 19, 2024, the companies announced a partnership, Dexcom invested $75 million in Oura’s Series D round, and they said glucose data would flow alongside sleep, stress, heart health, and activity data from the ring. (investors.dexcom.com) By 2025, that became a consumer feature inside the Oura app. Oura’s help pages now say United States users with an active membership can connect the Stelo glucose biosensor from Dexcom and view glucose data next to Oura insights about meals, activity, sleep, and stress. (support.ouraring.com) The jump from product feature to personal model is small in code and large in ambition. A company dashboard usually shows yesterday’s score, while a custom workflow can ask narrower questions like whether a bad night of sleep predicts a bigger breakfast spike, or whether a hard workout changes your afternoon glucose curve the next day. (cloud.ouraring.com) (support.ouraring.com) That is why this looked like a live demo of where wearables are heading. The hardware is already on people’s bodies, the application programming interfaces already exist, and the missing layer is software that turns scattered signals into a personal model that updates as fast as your own habits do. (cloud.ouraring.com) (investors.dexcom.com)