Tempus secures up to $9.5M from ARPA-H for an autonomous AI agent in heart failure care

Tempus has secured up to $9.5 million from ARPA-H to build an autonomous AI agent aimed at heart failure care, according to a Pulse 2.0 report. The item states the funding ceiling and the intended use, but does not describe the product form, the technical approach or the project timeline.

ARPA-H is the US federal agency for advanced health research, and its awards typically go to high-risk directions that still need engineering proof. Directing money at an autonomous agent suggests chronic, data-heavy care pathways such as heart failure are being treated as a suitable proving ground.

Autonomy here means more than answering questions: an agent is expected to push tasks forward within a defined mandate, for example tracking patient data, triggering follow-ups or coordinating steps in a care pathway. Once that reaches clinical settings, accountability, authorisation boundaries and auditability become preconditions rather than later additions.

Heart failure magnifies both the upside and the risk. Patients need long-term monitoring and frequent adjustments that span hospitals, clinics and the home, so any automation has to sit alongside existing clinical decision-making rather than replace a physician's final judgement.

The signal for the sector is where the money is going. Healthcare AI has clustered around clearly bounded, single-step tasks such as imaging, documentation and coding, while this award is framed as an autonomous agent for a named condition, which means a research funder rather than just vendors is betting on agent-form care tooling.

Three things to watch: the award's execution period and milestones; whether Tempus publishes a clinical validation plan and safety boundaries; and whether ARPA-H extends similar funding to other chronic conditions.