Ask most frontline clinicians what they need from AI, and the answer rarely starts with more data. The monitors already sound the alarm. The dashboards already flag. The EHR already surfaces risk scores. What clinicians need, and what most AI deployments still fail to deliver, is action.
That gap between insight and execution emerged as a defining theme at a panel on care model transformation at the Becker’s 16th Annual Meeting, where executives described what it actually takes to move AI from advisory to operational.
“The gap today is not insight,” said Jason Cohen, MD, chief medical officer of Qventus. “We don’t need more intelligence. As a clinician, please don’t give me another alert, but actually start automating the next step in the workflow for me.”
The example he offered was specific: a sepsis alert that stops at the notification is a half-built solution.
“I don’t need a sepsis alert,” Dr. Cohen said. “I need something that tells me when a patient looks like they may have sepsis. They don’t have a lactate ordered yet. Do you want me to order that for you? By the way, it looks like this chest X-ray this morning had a suggestion of a right lower lobe pneumonia.”
That distinction between surfacing a finding and completing a workflow step is where most health system AI implementations currently stall. Closing that gap requires more than a technology upgrade.
“That requires health systems actually starting to figure out how to basically build that muscle of moving from insights and actions to execution, which is where both the value is and where we’re going to see a reduction in burden for our frontline teams,” he said. “But that requires a lot of workforce development.”
That muscle has a specific anatomy: people who understand clinical workflows deeply enough to build against them. Dr. Cohen described the need for actual builders — staff with enough workflow fluency to translate AI output into reliable operational steps at the point of care, not just analysts who can run models.
Michelle Stansbury, associate chief innovation officer and vice president of applications at Houston Methodist, offered one of the panel’s most concrete examples of what full operationalization looks like: an AI voice agent that handles inbound patient scheduling calls end to end, with no human handoff required. The rollout required its own change management strategy.
“I’ve had to have people listen to that call recording because they don’t believe it,” Ms. Stansbury said. “But once they do, then they’re your change agents going forward.”
Converting skeptics through dynamic examples reflects a broader truth about operationalizing AI. Workflow transformation rarely succeeds through top-down mandate. It succeeds when frontline staff experience the change as relief rather than imposition.
At Jefferson Health, Judd Hollander, MD, senior vice president of healthcare delivery innovations, has institutionalized that principle through formal governance. The system created eight SWAT teams — an acronym Dr. Hollander coined for Synchronizing Workflows and Technology — each staffed with leads from virtual care, nursing and medical informatics, IS&T, and a project manager, operating on a standing biweekly cadence with required checklists and workflow documentation before any solution goes live.
“It really is all about the workflows and operations and not the technology,” Dr. Hollander said.
The structure keeps virtual care strategy tightly coupled to IT implementation, a connection that at many health systems exists in name only. Dr. Cohen identified a second, less-discussed prerequisite for execution: leadership teams with enough technical fluency to make accurate decisions about what to build, what to buy, and what to wait for. That means knowing which capabilities are genuinely differentiating versus which are already on the EHR roadmap, and resisting the pressure to buy solutions that address visible problems while pushing complexity elsewhere.
“The future is less going to be defined by what we actually build and more by how we build it and who we build it with,” Dr. Cohen said. “Those health systems that figure out how to turn intelligence into execution and actually weave intelligence into existing workflows and build those solutions are going to be the ones that are successful.”
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.
