PIA: A Personal Intelligence Agent Turning Health Conversations into Records and Records into Understanding

2026-09-25

AI Summary

This study addresses the limitations of general-purpose agent memory in healthcare, including clinical information loss, ambiguous temporal parsing, and inadequate long-term trend modeling, by developing a personal intelligent health agent that transforms conversations into structured clinical records. Methodologically, it proposes a synergistic architecture integrating a domain-agnostic memory framework with pluggable health modules, leveraging knowledge graphs, medical lexicons, and temporal rules for structured extraction. Through four core mechanisms—extraction, storage, retrieval, and comprehension—the system achieves a cognitive leap from one-dimensional recall to three-dimensional causal trajectory modeling. Experimental results demonstrate that the system effectively filters approximately one-third of structural noise, validates the impact of query phrasing on comprehension quality, and enables precise synthesis of longitudinal health trajectories.

Abstract

General-purpose agent memory summarizes conversations: it extracts salient snippets, embeds them, and retrieves the top-k into the prompt. A health agent cannot run on summaries: a dose becomes a sentence, "since last week" is resolved at the model's discretion, and a three-month glucose trend cannot be answered by text similarity. We present PIA, a personal intelligence agent deployed alongside a consumer health agent. PIA receives the agent's natural-language requests, decides for itself whether and how to write or read, and turns conversations into typed clinical records and records into a synthesized understanding of the user. Its memory harness consists of four controls -- extraction, memory, retrieval, and understanding -- each a domain-agnostic mechanism with a pluggable health module: schema, medical alias dictionary, knowledge graph, and temporal rules. We show how the same query receives a different answer as the memory injected into the response context deepens from one-dimensional recall, to a two-dimensional health snapshot, to a three-dimensional trajectory with causality, and report lessons from operation: self-reported health data are missing not at random, question phrasing governs the quality of synthesized understanding, and nearly a third of candidate causal links are structural noise that rules alone remove.

Innovation

• Personal Intelligence Agent
• Structured Clinical Records
• Memory Harness
• Knowledge Graph
• Causal Trajectory