Researchers develop open-source personal health agent using consumer data
A team of researchers has created a new open-source framework for a personal health agent that processes data from wearables and medical records.
System design and user needs
The developers analyzed web searches and forum queries to identify four specific user journeys. These findings led to the creation of three specialist sub-agents for different health tasks. The framework integrates data science capabilities with domain expert knowledge and coaching functions.
Data sources and evaluation
The agent reasons over multimodal inputs from consumer wellness devices alongside traditional medical records. Researchers evaluated the full system across eleven benchmark tasks using both automated tools and human review. The project involved more than 7,000 annotations and thousands of hours of work from experts.
Available framework details
The resulting system is an open-source multi-agent framework designed for dynamic health interactions. It allows users to receive personalized support based on their longitudinal data history.
Reported by one outlet
Only one outlet has published this. Nothing here has been checked against a second report, so read it as that outlet's account and follow the link below for the original.
Reported by
1 independent outlet. Headline as published. Links open the original report.