Key Takeaways
- Mohammad Noshad and Shyld AI are using peer-reviewed research to demonstrate how Physical AI can move beyond monitoring hospital environments to taking action in real time.
- Shyld AI’s edge-native approach allows its technology to understand activity inside physical spaces and respond without relying on hospital staff to initiate each action.
- Mohammad Noshad’s vision for Shyld AI extends from autonomous infection control into operating rooms, pharmaceutical facilities, and other complex environments where AI can understand workflows and act in the physical world.
Mohammad Noshad, co-founder and CEO of Shyld AI, is building hospital technology that does not wait for a person to act. His company makes devices that detect a contamination risk and respond immediately, without a clinician or an environmental services worker entering the room. Shyld AI’s hardware now runs in more than 30 U.S. hospitals, and the approach has drawn peer-reviewed research support along the way.
The problem behind that growth remains large. According to a CDC report published in July 2026, roughly 1 in 38 hospitalized patients now has a healthcare-associated infection, down from 1 in 31 in 2015, totaling an estimated 518,000 such infections across U.S. hospitals each year. Progress has been uneven. Surgical site infections and pneumonia cases have barely moved even as other categories improved.
Noshad’s interest in the problem began with a personal loss. A close friend went in for a routine surgical procedure and did not survive an infection contracted at the hospital. The experience led Noshad and his brother Morteza to found Shyld AI in 2022, with a plan to put safety tools directly inside the rooms where care happens rather than in a report generated after the fact.
Sensors and UV-C light replace a manual routine
According to Shyld AI, a unit mounts on a hospital wall and is similar in size to a smoke detector. Onboard sensors continuously monitor the room, looking for moments that increase contamination risk, such as a shared keyboard touched by more than one patient or a surface left uncovered during a room turnover. When the system flags one of those moments, it fires a targeted dose of UV-C light, a high-energy wavelength of ultraviolet light long used for disinfection, that clears the surface within seconds.
Most hospital AI products stop at the alert. A system identifies a problem, sends it to a dashboard, and waits for a staff member to read it and respond. Shyld AI built its hardware to close that loop on its own, carrying out the disinfection step the moment it detects a risk, with no one needing to start the cycle by hand.
The device runs on VERTEX, Shyld AI’s own foundation model, built in partnership with NVIDIA to process what it sees locally rather than send it through a hospital’s cloud systems. Local processing keeps a unit running even during a network outage, ensuring no video or patient data leaves the room. Hospital IT and compliance teams have found the setup easier to review than most new clinical software, since a device evaluated on that basis looks closer to a piece of building equipment than a system tied into hospital records.
What the Stanford data adds to Shyld AI’s pitch
A peer-reviewed 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, tracked the system for several weeks and found it reduced cumulative microbial bioburden by more than 93% compared with a room cleaned by hand.
That result has coincided with faster sales. Shyld AI’s typical sales cycle now closes in eight to ten weeks, compared with the twelve to eighteen months hospital technology purchases usually take. “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 said.
The company has also backed that momentum with capital. Shyld AI closed a $13.4 million seed round led by Aulis Capital in May 2026, with the funds it plans to use for new hospital installations and continued engineering work on VERTEX.
Two founders who point to one moment as the reason
The brothers bring different training to the company. Mohammad completed his PhD in two and a half years and spent several years in AI research at Harvard, having already founded and exited two companies before starting Shyld AI. Morteza holds a doctorate in computer science from Stanford and built the technical architecture behind VERTEX.
Their plans for the technology extend past hospital rooms. Inside operating rooms, Shyld AI’s devices already track surgical readiness and flag missing instruments before a case starts, work designed to catch delays before they push back the rest of the day’s schedule. The company has also begun working with pharmaceutical manufacturers to bring the same sensing and disinfection approach to cleanroom environments, where contamination control is equally important.
Noshad describes the goal as building an entirely new category: autonomous AI designed to act within physical spaces that most software only monitors. He still measures the company’s progress against the outcome he set out to prevent, tracking each new hospital deployment against the loss that started the company in the first place.
To learn more about Shyld AI’s technology or schedule a demo, visit shyld.ai or connect with Mohammad Noshad on LinkedIn.
About Shyld AI
Shyld AI is a healthcare technology company bringing physical agentic AI to hospital operations. Founded by CEO and Co-Founder Mohammad Noshad, Shyld AI develops autonomous physical agents that streamline hospital operations like infection control, OR efficiency, and compliance, without adding workload for staff. Its technology combines AI with UV disinfection to reduce environmental contamination by up to 93%. Shyld AI is deployed across hospitals nationwide, improving clinical care, efficiency, and cost savings. To learn more, visit www.shyld.ai/
