Epic Expands AI Ambitions With Agent Platform, Cosmos-Powered Predictions and Workflow Automation

The announcements were made at Epic’s annual Users Group Meeting (UGM) in Verona, Wisconsin, where CEO Judy Faulkner outlined the company’s broader AI strategy around “Ergo,” which Epic describes as healthcare intelligence combining human judgment with AI-driven analysis of large-scale health data.

Epic is expanding its healthcare AI strategy with a new platform for building AI agents, a predictive AI model trained on its Cosmos clinical database, and additional tools designed to automate clinical, operational, and administrative workflows.

The announcements were made at Epic’s annual Users Group Meeting (UGM) in Verona, Wisconsin, where CEO Judy Faulkner outlined the company’s broader AI strategy around “Ergo,” which Epic describes as healthcare intelligence combining human judgment with AI-driven analysis of large-scale health data.

Epic said Ergo is built on data from its electronic health record (EHR), MyChart, its healthcare network and Cosmos. The company is also positioning its AI tools, including Art for clinicians, Emmie for patients and Penny for revenue cycle operations, as components of an integrated enterprise AI system.

On the clinical side, Ergo Visit is now live at Ochsner Health and with five physicians participating in Epic’s Garden Plot collaborative. The workflow uses Art to summarize relevant patient records before visits, while Emmie interacts with patients through MyChart to identify potential topics for discussion. Epic said Art-generated summaries are currently in use at more than 300 healthcare organizations.

Epic also highlighted Agent Factory, a platform that allows healthcare organizations to build, deploy, and monitor AI agents capable of reasoning and taking action across workflows. The platform can be configured without coding, and Epic said early adopters are already using it. Agent Factory is expected to become widely available in 2027.

ECU Health, one early adopter, developed an AI agent to summarize transfer center requests. According to Epic, the tool improved handoffs and saved an estimated 20 hours per week.

Another major development is Curiosity, a generative AI model built using data from Cosmos. Epic said Cosmos includes clinical information from its participating health systems representing more than 310 million patients and 21 billion medical encounters.

Curiosity is being validated by 20 healthcare organizations and is designed to generate predictions such as hospital readmission and stroke risk. Epic said the model can address population-level clinical questions and provide insights into an individual patient’s potential health trajectory. The tool will be available to organizations contributing data to Cosmos.

Epic executives said the company is also focused on integrating these capabilities directly into EHR workflows as healthcare organizations and patients increasingly adopt AI tools.

Stay tuned for more such updates on Digital Health News