Closing the Loop: why we invested in Kai Health
The importance of diagnosis errors in American healthcare
American medicine has spent 25 years, and untold billions, making the medical record better. Electronic health records, dictation software, and ambient AI scribes made the note faster to produce and easier to read. Almost none of that investment went toward using the record to change the care a patient actually receives. The loop stays open.
One of the clearest causes is human error in diagnosis. In 1999, the Institute of Medicine’s To Err Is Human put medical error on the national agenda. Sixteen years later, the National Academy of Medicine concluded that most people will still experience at least one diagnostic error in their lifetime. It remains the single most common allegation in emergency department malpractice claims. A generation of spending made the notetaking AFTER diagnosis better. But all that spend did not improve the diagnosis.
And the moment of diagnosis is where a health system’s exposure is greatest. The standard response looks backward: outsourced chart audits and compliance reviews that surface gaps long after the patient has left, too late to reduce risk or repair the record.
The team
That gap is what Kai Health was built to close, and its founders have spent their careers on little else. In 1998, Dr. Dan Sullivan, an emergency physician with a law degree, founded The Sullivan Group , a patient safety and risk management firm with one goal: reducing medical error and the malpractice claims that follow. Its hallmark product is RSQ®, short for Risk, Safety, and Quality.
Dr. Sullivan’s conviction, formed in the country’s busiest emergency departments, was that preventable errors are failures of systems and tools, not of clinician skill. RSQ® turned that conviction into a patented methodology: chart review tied to clinical standards validated through independent research on roughly 170,000 high-risk emergency patients, with findings fed back to the individual physician. The largest hospital system in the country adopted it in its emergency departments in 2001 and reported a drop in diagnosis-related malpractice claims of more than 70%. RSQ® now runs in more than 1,000 acute care facilities and touches over 20 million patient visits a year. The clearest sign it works is who pays for it: malpractice insurers award premium discounts to clinicians who complete its programs.
Kai Health takes that methodology and makes it autonomous. Brant Roth, who leads Kai as CEO, is the other half of that history: fifteen years at The Sullivan Group, first selling RSQ alongside Dan, ultimately running the company as its President. That work made him fluent in the two markets that matter most: the carriers and captives that price medical malpractice risk, and the self-insured health systems that carry it on their own balance sheets.
Kai’s AI reads finished notes in real time, checks them against the RSQ indicators, and routes every gap that creates malpractice exposure or lost revenue to the physician, quality team, and executives who own the budget.
Put simply: Kai reads what a clinician documented, catches what is missing before the case closes, and turns each gap into measurable follow-up. It is that rare company: AI-native, with thirty years of proprietary clinical data and the trust that came with it.
That combination shows up where it is hardest to fake: the sales cycle. Selling software into large hospital systems is famously slow. But not for Kai.
In under a year it signed multi-year contracts with five customers, three large health systems and two physician groups, including one of the country’s largest non-profit systems. The reason is the founders. Three decades of relationships and credibility let Dan and Brant skip the procurement gauntlet and reach the executives who own clinical risk.
“I am beyond thrilled with all that is to come from our partnership. This was energizing to see how well the first one went and to appreciate the scale at which we will make a difference.”
Natural Director of Emergency Department Risk Management at one of the country’s largest health systems
Why now
Two forces make this the moment.
First, medical liability is getting more expensive, quickly. The most comprehensive estimate puts its cost at $55.6 billion a year, a 2008 figure worth more than $80 billion today. The trend is steep: the average malpractice claim is up roughly 50% since 2009, and medical professional liability is now the most severity-pressured line in casualty insurance. Diagnosis-related events in the emergency department top that curve. For systems that carry their own risk, it is one of their biggest and least-managed exposures, and the budget to fix it already sits with the malpractice carrier and the captive, not the operating line. When the party on the hook will pay to reduce its own exposure, you have a fast way in and a loud signal about the problem’s size.
Second, the wedge is large and uncontested. Clinical risk and diagnostic safety in the emergency department is a substantial market on its own, and no AI-native company occupies it today. From there the thesis widens: the same engine can measure quality across any service line, automate the chart audits hospitals still run by hand, and ultimately connect documentation to patient outcomes. Health systems handed the revenue cycle to Optum and R1 because it was too complex to run in-house; quality and risk operations are at the same inflection. Kai is built to be that layer.
Why Distributed Ventures
At Distributed Ventures, our focus is the future of risk. Through our advisory and strategic LP base, we saw firsthand how large medical liability has become for the country’s largest self-insured health systems. Kai maps onto that insight precisely: a proven way to take risk and cost out of a line item they have struggled to manage for decades. We led the company’s seed round, joined by strategic co-investors with deep roots in emergency medicine and revenue cycle management.
The bigger picture
We back companies that help individuals, families, and institutions manage risk better. Kai sits where clinical data becomes measurable risk reduction, and that point is about to matter a great deal. Systems that adopt it will not only lower malpractice exposure. They will close a loop that stayed open through every prior wave of healthcare technology. The automation that reshaped the revenue cycle is coming for quality and risk operations, and the company that connects documentation to outcomes will define the category.
If you are building at the intersection of clinical AI and risk, we would like to hear from you.
We welcome any thoughts or questions.
Antonio Calderon can be reached at antonio@distributedvc.com.