It's 6:47 a.m. in a hospital you've never been to.
A charge nurse stares at a whiteboard. The OR schedule has two cancellations and a hole big enough to drive a bus through. Three patients in the ICU could go home today, theoretically, if someone could just chase down a social worker, a pharmacy clearance, and a ride. The day has not started and it is already behind.
Somewhere in the background, a piece of software is watching all of this. It is predicting who will be ready to leave, nudging the case manager, filling the OR gap with a procedure that was on a waitlist three weeks ago. The software is called Qventus. The people using it would tell you it is the closest thing they have to a teammate who never sleeps and never forgets to follow up.
Hospitals don't need more dashboards. They need decisions.
The American hospital is brilliantly understaffed and beautifully broken.
The numbers, for the curious-but-skeptical reader, are unkind. U.S. health systems lose an estimated tens of billions of dollars a year to operational friction - delayed discharges, empty OR blocks, unfilled pre-admission testing slots, patients waiting in EDs for inpatient beds that exist on paper but not in practice. Nurses spend a frankly absurd share of their shift hunting for information that should have arrived in their inbox an hour ago.
The conventional fix, for the last twenty years, has been a dashboard. Add a Tableau. Hire a consultant. Send a weekly PDF to a vice president. None of this, it turns out, moves a single patient through the building any faster. The dashboard sees the problem. It does not fix it.
You can't make a hospital more efficient by emailing it a chart.
Above: roughly the moment the Qventus founders started taking notes.
Three operators walk into a McKinsey project.
The origin story is unfashionable in its lack of drama. Mudit Garg, then a Stanford-trained engineer working in McKinsey's healthcare practice, kept noticing the same thing in client after client - the data inside hospitals was rich, the decisions made on top of it were not. He pulled in two collaborators, Ian Christopher and Brent Newhouse, and in 2012 they started a company first called AnalyticsMD. The name was descriptive and, like most descriptive names, a little flat. It became Qventus a few years later.
Their bet was unfashionable too. While the rest of healthcare AI in the early 2010s chased radiology images and biomarkers and the seductive headline of "AI diagnoses cancer," Qventus pointed itself at the unglamorous middle of the hospital - operations. The hallways. The schedule. The handoff. The bet was that real value sat in the boring places, and that nurses and OR managers, given a tool that actually acted, would adopt it.
The boring middle of the hospital is where the money - and the suffering - actually live.
How Qventus got from a slide deck to 100+ hospitals.
It's not a dashboard. It's a teammate.
Qventus' framing in 2025 is "AI teammates" - software that lives inside the EHR, looks at what is happening in real time, and then does something. It nudges the case manager about a discharge barrier. It fills an empty OR block with a real patient. It tells a pre-op coordinator which surgery is about to fall apart and what to do about it. The aesthetic difference between Qventus and the dashboards of a decade ago is the difference between a smoke detector and a fire department.
Under the hood, the platform is a stack of machine learning models on hospital operational data, wrapped in workflow automation, wrapped in - this is the unusual ingredient - behavioral science. Qventus's team will tell you, with the patience of people who have explained this a thousand times, that prediction without behavior change is theater. The product is engineered for adoption first, intelligence second.
Surgical Growth
Finds open OR time, fills cancellations, and helps systems add high-value cases without adding rooms.
Perioperative Coordination
Automates pre-admission testing and prep so surgeries don't fall apart the day of.
Inpatient Capacity
Predicts discharges, surfaces barriers, shortens length of stay.
Malnutrition Care
Spots at-risk patients and automates nutrition consults and documentation.
Four products. One unifying idea: do the boring work so the humans can do the hard work.
The receipts, in numbers.
It is easy to claim impact in healthcare. It is harder to publish numbers a CFO will accept. Qventus has spent the better part of a decade collecting both. Reported outcomes from health-system deployments include sharp drops in surgical cancellations, double-digit reductions in excess hospital days, and measurable gains in staff productivity - the kind of metrics that, in aggregate, are how a hospital actually keeps its lights on.
Reported operational impact
Bars are illustrative. The KLAS score is not - 92.5 is rare air in hospital software.
The customer list, meanwhile, has gotten heavy. Banner Health. Allina Health. HonorHealth. Boston Medical Center. Jackson Health. Northwestern Medicine. NewYork-Presbyterian. The most telling signal arrived in January 2025, when Northwestern, HonorHealth and Allina did not just renew - they put venture capital into the Series D. Customers paying you and investing in you in the same week is not a coincidence. It is a thesis.
Three of our biggest customers wrote checks in the same round. We took the hint.
Simplify how healthcare operates.
Qventus's stated mission is straightforward to the point of being unfair to write about. Simplify how healthcare operates so care teams can focus on patients. There is no five-word vision statement involving "synergies." The company believes, in a way that is genuinely culturally embedded, that the work of running a hospital is honorable work that should be aggressively unburdened by software. Employees rotate into hospitals. They watch the OR schedule get torn up at 5 a.m. They write code that knows the difference between an algorithm and an actual nurse's morning.
Predicting the future is fine. Doing something about it is the entire job.
What this looks like tomorrow.
The interesting question is not whether AI will be inside the hospital. It is already inside the hospital. The interesting question is whether the AI will sit in a tab nobody opens, or whether it will earn a chair at the bed huddle. Qventus is making a long, patient, unglamorous bet on the second outcome. Agentic software. Workflow-first design. Behavioral nudges over PDFs. A platform that does not ask the user to log in and admire it.
If they are right, the next decade of hospital operations starts to look less like a heroic act of human coordination and more like a quietly competent partnership between care teams and software that just handles the rest. If they are wrong, the rest of the healthcare AI industry catches up by accident, because Qventus has spent ten years quietly proving the playbook.
It's 6:47 a.m. again. Same hospital. Different morning.
The whiteboard is mostly clean. The two OR cancellations from yesterday were backfilled overnight by a piece of software that called the patient, confirmed pre-op clearance, and put the case on the schedule before anyone had to ask. Three discharges are queued. Pharmacy already knows. The ride is booked. The charge nurse is making coffee.
The software is still watching. It is also, finally, doing something about what it sees.
That is the whole product.