Bunkerhill Health Secures $55M to Push Agentic AI Into Clinical Practice

The capital raise, backed by Sequoia Capital, Felicis, Optum Ventures, and Y Combinator, arrives as US healthcare spending reaches $5.3 trillion alongside persistent labor shortages. Bunkerhill’s platform deviates from off-the-shelf software, allowing hospitals to build custom agents for tasks ranging from cardiology imaging review to administrative prior authorizations. CEO Nishith Khandwala argues that medicine has advanced faster than the ability of clinical systems to operationalize new insights, leaving providers with more opportunities to improve outcomes than they have staff to execute.

Scaling Clinical Adoption

The University of Texas Medical Branch (UTMB) currently runs more than 20 agents on the platform. In one instance, a coronary calcium detection agent flagged a patient at imminent risk of a heart attack, leading to a successful triple bypass. Other deployments at UTMB include a nephrology triage tool that reportedly cut specialist wait times by over 50 percent, and a lung nodule tracker that accelerated urgent case responses by 80 percent. While these results underscore the potential of agentic systems, they remain self-reported metrics tied to specific institutional conditions rather than independent, multi-center trials.

Vinod Khosla, whose firm Khosla Ventures participated in the round, noted that the primary bottleneck in healthcare AI has historically been implementation rather than technical capability. As Bunkerhill expands, it must address the governance challenges inherent in allowing departments to build their own agents. With the new funding, the company plans to broaden its clinical use cases while bolstering the monitoring and liability frameworks necessary for health systems to trust autonomous agents with critical decision-making.