In# AsthmaWell

Agentic AI population health platform for asthma care coordination. Four specialized clinical tools and a production MCP server.

What AsthmaWell Does

About 1 in 12 Americans lives with asthma — roughly 26 million people. The burden is not evenly distributed: adults in high-pollution ZIP codes experience significantly more emergency visits. The data identifying who is at risk is largely available today, structured in FHIR R4 electronic records. A common problem is the gap between data and coordinated clinical action.

HEDIS quality measures exist to close that gap. The Asthma Medication Ratio and five related measures are tracked by payers, embedded in value-based contracts, and correlated with preventable hospitalizations. But a practice managing hundreds of asthma patients cannot review every chart weekly. The measures identify populations; they do not produce care plans.

AsthmaWell closes that gap with four specialized, LLM-backed tools. The pipeline assesses respiratory risk, evaluates six HEDIS-aligned quality gaps, synthesizes a structured care plan routed to the right roles and settings, and drafts outbound communication. Every output is framed as a structured signal for a licensed clinician's review — not a clinical order, not a diagnosis, not medical advice.tension specification shaped the server's initialize response design.


AsthmaWell does not provide medical advice, diagnoses, or clinical decisions. All outputs are structured signals for review by licensed healthcare professionals.

Built With

Share this project:

Updates