When UT Health San Antonio opened the UT Health Multispecialty and Research Hospital in late 2024, the facility wasn’t designed simply to expand the system’s inpatient capacity. It was designed as infrastructure, a physical proof point that artificial intelligence had moved beyond the pilot phase and into daily healthcare delivery.
Edward Sankary, MD, vice president, chief health informatics officer and chief value officer at UT Health San Antonio, and president of the regional physician network ACO, didn’t know how the technology implementation would go. But he knew the system had to move forward.
“A few years ago, most organizations were just trying to pilot things, pilot AI use cases,” said Dr. Sankary. “Now these tools are showing up throughout our entire workflow spectrum and really integrated into what we’re doing.”
Every patient room in the new hospital was built with a video platform and a wall-mounted camera, speaker and microphone, enabling virtual nurses to observe patients, verify medications, guide patients through the discharge process and respond to rapid response situations from a centralized location. Building the technology into construction was a deliberate choice since retrofitting it into an existing facility would have been far more difficult.
The hospital includes ambient documentation and three AI agents within the EHR allowing patients to schedule appointments, prepare for visits with a natural language conversation and a revenue cycle agent making suggestions on codes and developing denial appeal letters. The EHR also includes a clinical decision support tool for nurses and physicians.
The organization has also invested in internal AI building. Through a secure portal, any member of the UT Health San Antonio organization can access the system’s large language models in an environment designed to keep patient information within the institution’s walls. Staff can use the platform to build and deploy custom agents for specific workflow problems, an approach that has already produced measurable compliance improvements in clinical operations.
None of that infrastructure holds without governance, where many organizations underinvest.
“Governance is an important part of successful artificial intelligence,” he said. “As we look at the significant number of AI applications that we have, the ability to accurately determine the risk of that AI application, how that application was trained, if there’s a risk bias, if there’s a risk of performance decay over time. We’re in the process of implementing an AI monitoring system and management system that’s going to allow us to be sure as we use, deploy and evaluate artificial intelligence that it goes through a process that determines risk and we can continue to monitor it.”
Looking ahead, Dr. Sankary is tracking several developments he expects will close the gap further between what AI promises and what it delivers in clinical settings. High on the list is Epic’s work on a predictive medical events model, which is AI designed to anticipate what comes next in a patient’s care journey.
“This is going to begin to predict what the next test is, what the next potential illness is, to allow us not only to help reduce the consequences of some of these illnesses, but maybe to prevent them,” he said.
He is also tracking the partnership with the newly merged UT San Antonio. In September 2025, UT Health San Antonio and the University of Texas at San Antonio merged to form UT San Antonio, now the third-largest public research university in Texas by research expenditures. Through a collaboration with the newly formed College of Artificial Intelligence, Cyber, and Computing, the health system is pairing AI and computer science PhDs with clinicians — applying deep technical expertise to problems that require medical context to solve correctly.
“The technology is obviously extremely powerful, but along with it, we’ve got to be sure that the workforce evolves with the technology,” said Dr. Sankary. “We’ve got to continue to watch that and help direct it to ensure we’re providing the right education for our workforce and our staff to be sure they’re maximizing the capabilities of this platform.”
Technology that outpaces training pipelines loses most of what it promises. He sees the health system gathering more data and information to serve patients better and improve the patient experience. He believes patients will start wanting to interact more with their health data and play an active role in their own care.
“We’re going to be moving away from just a portal that they log into and search around to find information to something where they can ask questions to the AI platform, with conversational AI about their healthcare and get the right information that they need to ultimately help their movement through the system,” he said.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
