By using our site, you consent to cookies.

Fortune Exclusive: Bunkerhill Health raises $55M to put AI agents to work inside hospitals

Skip to content
Loading...

The distance between an idea and reality

Written By:

Nishith Khandwala

Almost every good idea in healthcare already exists. It is sitting in someone's head right now. The hard part was never thinking of a better way to care for a patient. The hard part is building it.

At Stanford, my cofounder and I built a solution that could read a routine CT scan and spot the warning signs of a heart attack, sometimes years before it arrived. It worked. And it was running almost nowhere. Not even at Stanford. The idea was right. The world just had no way to turn it into reality.

Then my father had a heart attack.

Here is the part I have never been able to shake. The warning signs were already there, sitting in scans he had taken years before. The algorithm we built was made to catch exactly those signs. Had it been running where he was cared for, it would have flagged him with time to spare. It wasn't running. So no one looked. The idea was right. The medicine was right. What was missing was any way to turn that idea into something that reached him in time.

Years later, at a health system running on the platform we built, that same idea saved a life. A police officer came in for a routine visit with no symptoms. Our AI agent read his imaging in the background, saw the coronary calcium, and made sure the risk was not left to chance. His care team brought him back in. He went into surgery for a triple bypass. The heart attack that was coming never came.

Same idea. Finally running. A life still here because of it.

Once one idea makes it into the real world, you see how many others are lined up behind it. A team knows exactly how a complex prior authorization should be assembled from the chart, so they build an AI agent that does it, and treatment starts on time. Someone knows which referral should jump the queue, so they build an AI agent that applies that judgment to every patient, and the sickest reach a specialist first. A clinician knows the set of preoperative steps that cannot be missed, so they build an AI agent that stays alert to every one, and no patient reaches surgery without them. None of these were missing ideas. They were good ideas waiting for a way to be built.

This is what AI agents are actually for. Not to replace the clinician. Not to be the hero. Health systems and the people inside them are the heroes, and they have never been short on ideas. Agents are the foundation underneath, the thing that finally lets a health system turn its own best thinking into care that reaches every patient, clinical and operational alike.

I have watched health system leaders build these agents in a single afternoon on our Carebricks platform. Care pathways. Specialty referrals. Inpatient safety. The paperwork that quietly decides whether care moves or waits. Even rare diseases. When building something new costs almost nothing, the only real limit left is imagination. And healthcare has always been full of people who imagine something better.

Picture what that adds up to. Not a health system with more staff or longer hours, but one where every good idea its people have actually gets built, and gets built for everyone. Risks caught early, while there is still time and still choice. Care that moves from reactive to proactive, from treating an illness to caring for a person long before that illness becomes an emergency.

That is why we built Carebricks, Bunkerhill Health's agentic AI platform, and why this next chapter matters so much to us. Our job is the hard, unglamorous work underneath, the clinical, regulatory, and engineering effort, so that the best idea in the room never has to die in the room again.

The ideas were always there. Now we can turn them into reality.