Here is a fact that sounds like a typo but is not: in the United States, only about one in five neurosurgeries is monitored for nerve damage while it is happening. Not because monitoring does not work. It works fine. The problem is arithmetic. Intraoperative neurophysiological monitoring - IONM, if you want to sound like you belong in the room - requires a trained neurophysiologist watching a live feed of electrical signals traveling through a patient's nervous system, ready to shout "stop" the instant a waveform does something it should not. There are not enough of those people. There are increasingly not enough of those people. And the number of spine and neurosurgical procedures that would benefit from them is projected to climb roughly 40% over the next decade.
When demand rises and the supply of a scarce human skill does not, you get one of two outcomes. Either the care gets worse, quietly, in ways nobody puts on a chart. Or somebody builds a tool. Nervio is trying to be the tool.
The company was founded in 2020 by two brothers, which is the kind of detail that usually gets sanded off a pitch deck but is actually the whole thing. Dr. Omer Zarchi runs an intraoperative neurophysiology service at Rabin Medical Center in Israel - he is the person who has sat in the operating room, for years, doing the exact job Nervio wants to augment. Nir Zarchi is the operator, the one who has built companies before. One brother felt the problem in his hands. The other knew how to turn a frustration into a corporation. This is a better origin story than most, precisely because it did not start with a market-size slide.
What AIM actually does
The product is called AIM, and its job description is refreshingly boring, which in medical AI is a compliment. During surgery, monitoring equipment is already generating a flood of raw electrical signals - evoked potentials, EMG traces, the electrophysiological weather of a nervous system under stress. Historically a human expert reads that weather. AIM ingests the same raw data and, using machine-learning models trained on prior cases, converts it into real-time alerts and recommendations. When something drifts toward danger, it surfaces the change quickly and presents it in a simple visual display, so the clinician spends less time squinting at squiggles and more time deciding.
Crucially, Nervio positions AIM as an adjunct - a layer on top of existing monitoring, not a replacement for the neurophysiologist. This is partly regulatory prudence and partly the honest shape of the bet. The near-term promise is not "fire the expert." It is "let one expert safely cover more operating rooms at once," which, given the shortage, is the version of the future that actually helps. Prof. Jay Shils, the company's director of clinical affairs and a three-decade IONM veteran, frames it as turning neuromonitoring into a new standard of care by letting experts monitor multiple cases. That is a scaling argument dressed as a safety argument. It is also, probably, both.
The unglamorous discipline of going slow
The most interesting thing about Nervio is not its algorithm. It is its patience. Before asking the FDA for anything, the company ran AIM across more than 1,000 surgical cases and built it with input from dozens of practicing neurophysiologists. In a category where the demo is easy and the trust is not, that ordering matters. Anyone can build a model that lights up on a slide deck. Convincing a surgeon to let it advise during an operation is a different and much slower business.
Nervio has also stacked its bench with people who know exactly how slow. US CEO Dr. Richard Vogel, appointed in December 2024, is a past president of the American Society of Neurophysiological Monitoring - not a growth-hacker parachuted in from another industry, but someone the field already respects. The regulatory affairs lead spent 13 years at the FDA before joining a startup, which tells you Nervio understands that in medical devices, the regulator is not an obstacle at the end of the process. The regulator is a design constraint from day one.
The economics hiding inside the safety pitch
It is worth being clear-eyed about why this is a business and not just a good deed. IONM sits inside a global market estimated at around $3.3 billion, with more than 700,000 monitored surgeries a year in the United States alone. That number is large, but the more interesting figure is the one next to it: the roughly 80% of neurosurgeries that go unmonitored because there is no expert available to do it. A company that sells only to hospitals already doing full monitoring is fighting over a slice. A company that can help extend monitoring into cases that currently get none is expanding the pie. Nervio is explicitly aiming at the second, harder, larger target - which is also why the "one expert, many rooms" framing keeps recurring in everything the company says.
The model, at least in outline, is business-to-business. AIM is sold into hospitals, spine and neurosurgery programs, and the IONM service providers - firms like Bromedicon and Medsurant Health - that already supply monitoring coverage to operating rooms. Nervio's advantage is that it does not ask any of them to rip out their existing equipment. AIM reads the raw signals those systems already produce and adds a decision-support layer on top. That is a deliberately low-friction way into a conservative buyer. Hospitals do not adopt new things in the operating room quickly, and "keep your workflow, add an assistant" is a much easier sentence to say to a surgical director than "replace your stack."
The competition is inertia
Nervio's real competitor is not another AI startup. It is the way things are done now: legacy monitoring hardware from established players, plus the outsourced staffing firms that send a human expert - remote or on-site - to watch each case. Those incumbents are entrenched, reimbursed, and trusted, and none of that is nothing. But the field's own experts concede it has seen little genuine innovation in over 20 years, and the labor model has a ceiling that arithmetic keeps pushing against. You cannot outsource your way out of a shortage of the very people you are outsourcing. That structural crack is the opening Nervio is trying to widen, and it is a more durable moat than any single algorithm, because the shortage is not going away.
Two hemispheres, one company
Nervio is structurally split, on purpose. Research and engineering sit in Israel - the founding is entangled with eHealth Ventures and Clalit Health Services, and the company has taken grant money from the Israel Innovation Authority. Commercial operations and regulatory strategy run out of Nashville, Tennessee, close to the US hospital market it needs to sell into. About 20 people hold the two halves together. It is a common medtech shape - Israeli R&D, American go-to-market - but it works here because the actual problem, the operating room, is the same on both continents.
The money, so far, is modest and deliberate. Nervio closed a $2.5 million seed round led by eHealth Ventures, announced in early 2024, and has been working toward a larger post-seed raise to fund regulatory milestones and early commercial deployment. This is not a company trying to blitz-scale. It is a company trying to earn its way into the one room where mistakes are permanent.
Why it is worth watching
The skeptical read on medical AI is that everyone promises augmentation and nobody ships it. Nervio's answer is that augmentation is not a tagline here, it is the literal product: one neurophysiologist, more cases, an AI flagging what needs a human eye. If it works - if regulators clear it, if surgeons trust it, if hospitals pay for it - then a scarce, expensive form of surgical safety becomes something you can extend rather than ration. If it does not, it will be because trust in the operating room is the hardest currency there is to earn. Either way, the problem is real, the founders lived it, and the field has not had a genuine update in twenty years. That is a rare combination, and it is why Nervio is a name to keep in the margins of your notebook.