Five finalists have been chosen from Rochester, Minn.-based Mayo Clinic’s nurse-led AI agent competition field and are now moving into proof-of-concept development
Mayo received more than 450 submissions to the competition, which challenged nurses to design largely autonomous AI solutions for problems they encounter in their own work. That field was narrowed to 33 semifinalists, and Mayo nurses then voted on five finalists to move forward. Those five concepts were showcased Aug. 26 at Mayo’s third annual Nursing Nexus event in Rochester, an internal gathering where nursing staff pitch and refine ideas for improving patient care and the workplace.
Finalist Anna Callahan, BSN, RN, nurse team lead at Jacksonville-based Mayo Clinic Florida, created an agent designed to speed up FMLA and short-term disability paperwork, a process she said currently requires multiple handoffs and a rotating set of staff responsible for completing it.
“This is something that is a pain point for our patients, as well as for us,” Ms. Callahan told Becker’s. “It creates a lot of administrative burden. So my idea was essentially: I want an AI agent to help me complete that documentation, and then a human would review that process for accuracy, and then we would have it signed by the physician so that that process gets cut from several days to weeks to potentially just hours if we can formulate it in the right way.”
Another finalist, Duc Chu, RN, a critical care nurse and quality leader at Mayo in Rochester, is developing an agent aimed at cutting down the time nurses spend searching for information they need in the moment — patient-specific guidance tailored to clinical setting and case type. He said the goal eventually is to move beyond a retrieval tool into something closer to clinical workflow support.
Both nurses described their projects as being at a similar stage: past the initial concept and into proof-of-concept development, with more building and testing ahead before any decision on wider rollout.
“We’ve shown that it can work, but will it work?” Ms. Callahan said. “Can we make it happen?”
Added Mr. Chu: “Any idea also has multiple analysis steps, and so we’re analyzing what the sources of the information are, how we can leverage those sources, and how we feed that into a machine learning or an agentic workflow to help solve this problem.”
The finalists are working with Microsoft to build out their agents. But neither Ms. Callahan nor Mr. Chu could point to a target pilot or go-live date. Mayo Clinic Chief Nursing Officer Ryannon Frederick, MS, RN, said that is by design: The health system wants to avoid pushing a tool into use before it’s both reliable and genuinely useful to the nurses who would rely on it.
“We just don’t add it in to add it in for technology’s sake,” Ms. Frederick said, adding that Mayo is “continuing to build it out till we get to that point, so that we have confidence in it, but also so it can assist enough to make it worth the time.”
Ms. Frederick said some of the ideas that have come out of Nursing Nexus and the broader agent competition — particularly tools focused on clerical work rather than direct patient interaction — are likely to reach practice sooner than tools that touch patient care directly, which require more validation.
“But I have the utmost confidence that the ideas that we heard about today are going to continue to move forward and really spread across Mayo Clinic,” she 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.
