KEY TAKEAWAYS:

  • Mohammad Noshad built Shyld AI around a simple gap in hospital AI: an alert does not solve a problem until someone has time to act on it.
  • Shyld AI runs decisions directly on the device, allowing its system to respond in real time without sending patient data or video to the cloud.
  • A peer-reviewed Stanford study found that Shyld AI’s technology reduced cumulative microbial bioburden by more than 93%, giving hospital leaders a measurable result to weigh against the cost of adoption.

American hospitals have made real gains from artificial intelligence over the past decade. Infection surveillance software now catches outbreak patterns that once took weeks to piece together by hand, and predictive models flag rising risk before it spreads across a ward. The open question for hospital leaders is what happens in the stretch after that alert fires, and healthcare-associated infections still contribute to roughly 72,000 U.S. deaths each year, according to the most recent CDC data.

Mohammad Noshad, founder and CEO of Shyld AI, has built his company around that stretch. Before entering healthcare, he earned his PhD in 2.5 years, spent several years as an AI researcher at Harvard, and built and exited two other technology companies. What turned his attention to hospitals was a close friend’s death from an infection contracted after a routine surgery, a loss that led him and his brother Morteza to found Shyld AI in 2022.

The Problem Noshad Set Out to Solve

Noshad’s read on hospital infection control comes from watching how much of it still depends on a person doing the same task correctly, every time, under pressure. “When a human is involved during cleaning processes and procedures, they are prone to making errors,” he told Healthcare Brew. “We saw an opportunity to bring AI into that space to make it much more efficient and basically automate everything so that you have consistency and efficiency in applying day-to-day disinfection in those hospital areas.”

He’s expanded on that point elsewhere. “With manual, there’s no way for you to monitor if these processes are being done properly,” Noshad told MedCity News in May 2026. “There’s a good chance that people are missing areas or the contact time of the chemicals is not enough.” Digital monitoring tools have made hospitals better at spotting these gaps. They haven’t removed the wait between spotting a gap and someone becoming free to close it.

Building AI That Finishes the Job

Noshad’s answer is a device that skips the alert entirely and handles the response itself. Shyld AI’s wall-mounted units combine sensors with UV-C disinfection, reading a hospital room continuously and firing a targeted dose of light the moment they detect a contamination risk, a shared keyboard, or an exposed surface between patients, without anyone stepping inside.

“For the first time, we’re building an AI that’s automating that,” Noshad has said. “It’s taking action instead of people going into the rooms and doing disinfection.”

Every decision runs on VERTEX, Shyld AI’s own foundation model, processed directly on the device rather than sent out to a hospital’s cloud systems. Morteza Noshad, Mohammad’s brother and Shyld AI’s co-founder, holds a PhD in computer science from Stanford and built the technical architecture VERTEX runs on, giving the company deep engineering ownership of the layer that enables real-time, on-device decisions. That local processing means a contamination event gets handled in seconds, and no patient data or video ever leaves the room. Noshad treats that as a requirement, not a nice-to-have, since a hospital compliance team reviewing a closed system has far less to scrutinize than one reviewing a cloud-connected vendor. Deployment has followed accordingly. Shyld AI now runs in more than 30 U.S. hospitals, and sales cycles close in eight to ten weeks, compared with the twelve to eighteen months that hospital technology purchases usually take.

What Comes Next for Noshad and Shyld AI

The clinical case behind all of this has been through formal peer review. A study conducted at Stanford Hospital’s Advanced Endoscopy Unit, co-authored by Noshad alongside Stanford researchers Monique T. Barakat and Timothy Angelotti, and published in the American Journal of Infection Control, found that Shyld AI’s system reduced cumulative microbial bioburden by more than 93% compared with a control room.

“We’re moving the industry from passive AI to Active AI, technology that understands how hospitals operate and improves workflows in real time without adding burden to clinical teams,” Noshad has said.

Noshad’s ambition for the company extends well past disinfection. Inside operating rooms, the same architecture already tracks surgical readiness, monitors workflow phases, and catches missing instruments before a case begins, addressing delays that carry real cost in a setting where every minute of OR time runs into the hundreds of dollars. Outside the hospital, Shyld AI has started working with pharmaceutical partners to bring the same approach into cleanroom environments, where contamination control runs under many of the same constraints as a hospital room.

Noshad sees hospital AI moving toward systems that sense a problem and resolve it in a single step, rather than routing it to another person first. What began as his response to one friend’s death now runs across dozens of U.S. hospitals, working every day to keep that same outcome from happening to someone else’s.