Task oriented agents improve complex wearable health data analysis
A Task-Oriented Multi-Agent Framework for Complex Wearable Health Analysis
Multiagent Systems
Summary
Wearable health devices produce complex data that often requires multiple steps to answer health questions, including retrieving data, analyzing it over time, and giving advice. The authors propose splitting these complex questions into smaller tasks handled by specialized computer agents, each focusing on one part of the problem. Their system keeps track of the tasks separately to make sure every part is done properly and the reasoning is clear. Testing on fake but realistic data shows their approach answers questions accurately with less processing effort and improves clarity and trust in the responses.
What this means in practice
- •For health app developers: Create apps that break health queries into smaller steps handled by specialized components to improve efficiency and clarity in wearable data analysis.
- •For fitness data analysts: Use distinct task agents to better separate data retrieval and analysis steps when working with multiple user health intents for clearer insights.
Authors
Kunpeng Yang
Abstract
Wearable health questions often combine data retrieval, longitudinal analysis, and health advice over structured records. Prompting a single large language model with a complete record and a composite query obscures whether every request is executed and which evidence supports the answer. We propose a task-oriented multi-agent framework that represents a composite query as distinct intents and typed tasks with explicit intra-intent dependencies. Specialized agents execute retrieval, analysis, and advice tasks; isolated intent states preserve request boundaries and evidence relationships before aggregation. We evaluate the framework on a synthetic dataset of $10{,}000$ virtual users with one month of longitudinal wearable records, covering structured data retrieval, multi-intent recognition, and overall response quality. Across $1{,}500$ retrieval questions, the Query Agent achieves $98.3\%$ accuracy, compared with $97.9\%$ for the Direct LLM baseline, while reducing average query-stage token consumption from $6{,}869$ to $3{,}136$. On $180$ multi-intent questions, the Manager Agent achieves $100.0\%$ Multi-Intent Coverage and $94.4\%$ Multiset Jaccard Similarity. Under the current synthetic evaluation setting, our method receives higher mean Trustworthiness and Transparency scores on both question categories, whereas Actionability does not improve consistently. These results provide preliminary evidence that explicit task organization can support task-relevant data access and data-grounded longitudinal analysis, while leaving health advice generation and validation on real wearable data as open challenges.