Serverless by design
Traditional “always-on” servers use energy even when nobody is using them. Our backend is built on an event-driven, serverless architecture using Google Firebase. Our computing resources scale to match how many people are using the app, and drop to near zero when idle, which avoids most of the energy waste associated with traditional always-on servers.
Low-carbon cloud infrastructure
Our platform is hosted on Google Cloud. Google reports that it matches 100% of its annual electricity use with renewable energy purchases, that its data centres run on around 65% carbon-free energy on an hourly basis, and that it aims to run on 24/7 carbon-free energy by 2030. Google also reports diverting 88% of operational waste from its data centres away from landfill.
Efficient AI
Using AI responsibly means using it efficiently. Our AI features use Google’s Gemini models, which run on Google’s custom Tensor Processing Units (TPUs), designed for performance per watt. Google reports that the carbon footprint of a typical Gemini text request fell by a factor of 44 over a year, and that a typical request uses about 0.24 watt-hours, less energy than watching television for nine seconds.
We only call AI when you ask the app to analyse something, such as a meal photo, so nothing runs in the background without a reason.
Continuous measurement
We see sustainability as part of good engineering. We use Google Cloud’s Carbon Footprint reporting to follow the emissions from our cloud use, so that as more people use Health Goals, our efficiency keeps pace.
The figures on this page are Google’s own published figures, not our measurements.
