DeLorean Artificial Intelligence

Somewhere inside a health plan's data warehouse, a patient who feels perfectly fine is quietly being flagged. Not by a nurse. Not by a lab result. By a piece of software named MARTI that has read this person's records and concluded, with uncomfortable confidence, that a hospital stay is coming - and that there is still time to stop it. That is a normal Tuesday at DeLorean Artificial Intelligence.

DeLorean AI is a Palm Beach, Florida company that sells one strange and specific promise: it will tell you what is going to happen to a patient before it happens, and then tell you what to do about it. The company calls this "predictive and interventive" intelligence. Most of healthcare calls it wishful thinking. DeLorean has spent six years trying to move it from one column to the other.

The pitch is built on a small act of branding mischief. The company is named after the car from Back to the Future. Its flagship medical AI is named MARTI, which is not a coincidence if you have ever met a man named Marty McFly. Cute? Sure. But underneath the wink is a serious claim: that the same records hospitals already collect contain a readable signal of the future, and that nobody was bothering to read it.

01 / THE PROBLEMThe system waits for you to get sick

Medicine is reactive by design. DeLorean's bet is that it doesn't have to be.

Here is the uncomfortable truth DeLorean was built around. Most of healthcare is reactive. A patient deteriorates, shows up in an emergency room, and the machine of modern medicine springs impressively into action - expensively, often too late. Chronic disease is the worst version of this. Kidney disease, diabetes, heart failure and COPD rarely arrive overnight. They build for years, leaving a trail in the medical record that nobody has the time or tools to follow.

The cost of that blind spot is not abstract. A small fraction of patients - the sickest one percent - drive a wildly disproportionate share of spending. Find them early, intervene, and the savings are enormous. Miss them, and they show up later as a crisis with a six-figure invoice attached. The problem was never that the data didn't exist. It's that reading it at scale, accurately, was somebody else's someday.

02 / THE BETThe founder who walked away from the comfortable job

A career spent building AI for giants, redirected at a single problem.

Dr. Severence MacLaughlin had the kind of resume that makes leaving it look slightly insane. He had served as Chief of Intelligence for the Western Hemisphere at Capgemini Invent and as Global Head of Artificial Intelligence for Healthcare and Life Sciences at Cognizant. He had built AI for some of the largest companies on earth. In 2019, he left to build it for one idea instead.

He bootstrapped DeLorean with his own savings. For years there was no institutional money - just a conviction that prediction in healthcare was solvable and a willingness to fund the proof himself. The first target was diabetes. The approach was deliberately unglamorous: take the messy, structured and unstructured data that organizations already own, and turn it into a mathematical profile of a patient's likely journey.

Exhibit A in the case for quitting a very good job: one founder, several years of personal runway, and a company named after a time machine.

03 / THE PRODUCTMeet MARTI, the AI with a name and an opinion

Not a diagnosis tool. A forecast, plus a recommended next move.

DeLorean's flagship is Medical AI, and its working name is MARTI. It does two things that, taken together, are unusual. First, it predicts - it reads a patient's records and estimates the probability of a future event, like a hospitalization or the onset of a disease. Second, and this is the part most predictive tools skip, it recommends the "next best action" a care manager or physician should take to change that outcome.

The platform organizes its work into disease modules: chronic kidney disease, diabetes, cardiovascular disease, behavioral health and depression, and respiratory disease including COPD. It can flag undiagnosed patients - people who have a condition that no one has named yet - and track how a disease is likely to progress. There is a second engine, too: a CRM AI that applies the same prediction-then-recommendation logic to sales pipelines, reportedly lifting client sales by around seven percent.

04 / THE PROOFThe numbers DeLorean keeps repeating

Skeptics ask for evidence. Here is what the company puts on the table.

A claim about the future is only as good as the receipts. DeLorean's headline figure is accuracy: 85 to 87 percent on disease prediction drawn from electronic medical, healthcare and hospitalization records. More striking is the validation story - the company describes its Medical AI as the first to be "biologically validated" by independent third parties, with predictions reported accurate at a 95 to 99 percent confidence level by three separate organizations.

Then there is the money, which is the language payers actually speak. DeLorean reports clients cutting roughly $100 per member per month on the sickest one percent of their population - about $1,200 a year, per person - and around $10 per member per month across everyone else. Hospitals using the system reportedly saw adverse outcomes fall 27 percent within the first 90 days. One publicly cited payer engagement involved records on some 35 million patients.

A scoreboard, not a halo. The interesting thing about DeLorean's numbers isn't that they're large - it's that the company invited outsiders to check them.

05 / THE MISSIONTrying to make prediction the default

Value-based care, but with the value found before the crisis.

DeLorean frames its purpose around value-based care - the slow industry shift from paying for procedures to paying for outcomes. The company's version adds a verb: prevent. It argues that providers and payers who can predict disease will treat that ability as a competitive necessity, and that those who can't will, in its blunt phrasing, become obsolete within five years. It is the kind of prediction a prediction company would make.

The customers reflect the ambition. Beyond health plans and hospitals, DeLorean has applied its engine outside medicine - consumer-goods giant Grupo Bimbo has been cited as a client for the sales side. The team, around 25 people spread across South Florida, Texas and Tennessee, is small for the size of the claim. That gap between headcount and ambition is either the risk or the whole point, depending on how the next funding round goes.

06 / TOMORROWWhy the flag in the warehouse matters

Back to the patient who feels fine.

Return to that data warehouse. The patient who feels fine is still flagged. The difference DeLorean is selling is what happens next. In the old world, nothing happens - the signal sits unread until it becomes a symptom, then a bill. In the world DeLorean is building, a care manager gets a name, a probability, and a recommended action while there is still room to act. The hospital stay that was coming, sometimes, simply doesn't.

That is the entire wager. Not that AI can diagnose - plenty of tools claim that - but that it can see far enough ahead, accurately enough, that the future becomes negotiable. Whether DeLorean can prove that across the whole messy reality of medicine is unsettled. But the company has done the one thing most prediction startups never do: it handed its forecasts to outsiders and asked them to check whether the future arrived on schedule. So far, it says, the future did.

Profile compiled from public sources including DeLorean AI's website, Refresh Miami, citybiz, Under30CEO and The SaaS News.
Figures are company-reported unless otherwise noted. Last reviewed June 2026.