The hard part of medical AI was never the model. It was getting a hospital to actually run it. Bunkerhill Health has raised fresh money to close that gap.

The startup closed a $25m Series B led by Khosla Ventures, it told Fortune in an exclusive. That takes its total funding to $55m. Sequoia, Felicis, Optum Ventures, and Y Combinator all joined.

The traction is real. Over the past year the company grew revenue 20-fold and signed more than a dozen health systems, the founder said.

What Bunkerhill builds

Bunkerhill sells AI agents through a platform called Carebricks. A hospital brings it a problem, from long wait times to missed follow-ups to paperwork, then turns its own idea into an agent that does the work.

The platform runs at 15 health systems, including Cleveland Clinic, Mayo Clinic, and Intermountain Health, the company said. It carries nine FDA-cleared clinical algorithms, one that spots silent heart-valve disease, another that flags osteoporosis risk.

The 💜 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!

A father’s heart attack

The idea is personal. Chief executive Nishith Khandwala started the work at Stanford in 2017, then nearly gave up when hospitals waved him off. In 2020, his father had a heart attack.

A cardiologist later told the family an old scan had already shown the risk. “I think we could have caught this earlier,” Khandwala recalled him saying. He co-founded Bunkerhill with David Eng, and the pitch has not changed since.

“Medicine has advanced faster than our healthcare system’s ability to operationalize it,” Khandwala said. “Why should a hospital need to work with 100 different companies to solve 100 different problems?”

The results, and the caveats

At the University of Texas Medical Branch, 22 agents are now live. In their first month, one that reads coronary scans flagged a patient at imminent risk. Doctors sent him to cardiology for a triple bypass, and his team credits it with saving his life.

Other agents cut nephrology wait times by more than half, and chased up lung findings 80% faster. UTMB’s AI chief Peter McCaffrey calls the platform “a shared brain” for the health system. “We don’t need superintelligence to solve our biggest problems,” he said. “We need average intelligence.”

The field carries real risks. Accuracy and privacy questions dog it, and Mayo Clinic now faces a lawsuit from a former research director over its AI oversight. McCaffrey says the doctor-patient bond must stay “sacrosanct.”

A crowded, well-funded field

Money is pouring into healthcare AI. Bunkerhill lands in the same wave as Neko Health and a run of medical-AI raises. Its backer knows the terrain: Vinod Khosla was an early OpenAI investor and has funded health startups for decades.

“The bottleneck in healthcare AI was never the technology, it was getting a health system to actually run it,” Khosla said. “Bunkerhill closed that gap.”

Sequoia’s Alfred Lin, who led the seed round in 2023, is relaxed about the crowd. “I much prefer having 1,000 flowers bloom,” he said, “and the best ones will become durable over time.”