Epic's Seth Howard on AI strategy and leadership after Rana exit

Epic Systems is moving forward with a distributed leadership structure following the departure of President Sumit Rana, who stepped down last month after 28 years with the company. Rather than naming a single successor, Epic assigned Rana's responsibilities to four executives: Seth Howard as executive vice president of research and development, along with senior vice presidents Mark Lipsky, Erv Walter, and Garrett Adams.

In an interview with Fierce Healthcare at Epic's annual Users Group Meeting in Verona, Wisconsin, Howard explained that the approach reflects the company's philosophy on career development and succession planning.

Why it matters

Epic's leadership transition comes as the EHR vendor positions itself as healthcare's operating system, not just its record keeper. How the company manages internal succession while accelerating AI development will determine whether it can maintain its market dominance against nimbler startups and cloud-native competitors. The distributed leadership model also signals Epic's commitment to promoting from within rather than bringing in outside executives—a strategy that could preserve institutional knowledge but may limit fresh perspectives.

A deliberate approach to leadership

Howard, a 19-year Epic veteran, said the company intentionally avoids predefined career paths, allowing employees to build unique roles based on their strengths and organizational needs. He noted that Epic celebrates numerous 20-, 25-, and 30-year employee anniversaries at monthly staff meetings, reflecting low executive turnover.

"Epic would never" hire an external search firm to fill senior roles, Howard said, referencing CEO Judy Faulkner's occasional joke about the practice. The company focuses exclusively on developing leaders internally.

The leadership changes have been accompanied by other high-level departures at Epic, though Howard did not address those specifically.

AI as operational infrastructure

At the Users Group Meeting, Epic emphasized that its AI strategy represents a continuation of decades-long development rather than a reactive pivot. The company announced several AI capabilities centered on workflow automation for patient engagement, workforce management, payments, research, and clinical decision support.

Howard highlighted Epic's unified data platform as a competitive advantage. Because Epic builds rather than acquires software, patient data remains consistent across applications—one date of birth, one allergy list, one medication record. This integration provides a foundation for AI agents that can reason and act across workflows.

Agent Factory and customer innovation

Epic's Agent Factory, previewed at HIMSS26 and rolled out this week, enables health systems to build, run, and monitor their own AI agents. The platform reflects growing demand from providers who want to customize AI tools for specific workflows and clinical protocols.

An Agent Factory build-a-thon held earlier this summer generated more than 1,000 ideas from customers. Howard expects health systems to create and share agents across Epic's community, accelerating collective innovation.

Cosmos Curiosity predicts patient trajectories

Epic plans to release Cosmos Curiosity in March 2027 as an integrated EHR component. Unlike single-purpose predictive models, Curiosity uses generative AI trained on longitudinal patient records from Epic's Cosmos database—representing over 320 million patients and 23 billion medical encounters—to forecast the next event in a patient's clinical journey.

The model can predict not just whether a patient will be readmitted, but why, along with the likelihood of specific diagnoses. This capability could function as personalized clinical decision support at the point of care.

Changing roles for IT professionals

As AI automates routine coding and configuration work, Epic sees IT roles evolving toward higher-value activities. Analysts and developers will spend more time partnering with clinicians to identify problems, design workflows, and build custom AI tools.

Howard's own background illustrates this approach. As a software developer on Epic's anesthesia team, he spent significant time in operating rooms observing workflows before designing software. Since last year's Users Group Meeting, Epic's R&D teams have spent 23,000 days on clinical immersion trips.

Epic's message to providers is that AI has moved from experimental to operationally viable, with practical applications across workforce, operations, and clinical care. Details were first reported by Fierce Healthcare.