Google builds a three-agent personal health assistant on Gemini 2.0
Start here
A Google research team has developed a personal health agent (PHA) using a collaborative architecture of three specialist sub-agents: one for data science, one acting as a domain expert, and one serving as a health coach. The system is built on Gemini 2.0 and was evaluated in the WEAR-ME study across 11 benchmark tasks with 1,165 participants, drawing on more than 7,000 human annotations and roughly 1,350 hours of evaluation. The architecture is notable for splitting the work of personal health guidance into complementary specialist roles rather than relying on a single model. What remains open is how the agent performs outside benchmark settings, including real-world clinical deployment, regulatory questions, and integration with patient care workflows.
Written by a model from the reports · updated 3 hours ago
Coverage timeline
Follow the coverage to see the different angles.
- NatureThe anatomy of a personal health agent
A Google research team built a personal health agent (PHA) using a modular multi-agent architecture with three specialist sub-agents: data science, domain expert, and health coach. Built on the Gemini 2.0 model, it was evaluated on 11 benchmark tasks in the WEAR-ME study (n=1,165), involving more than 7,000 human annotations and 1,350 hours of evaluation.
Coverage and heat for this story
Reports so far 2·Heat now 6·Peak heat 10 (Oct 21 08:00)·Reports, 24h +2
Two panels, same clock (pointer stays in sync): top is cumulative editorial reports — only rises; last point = All reports; slope = new reports that hour. Bottom is Live-board heatIndex — 7-day participants with a 24-hour half-life, so it can fall when talk cools.