Healthcare organizations are finding use cases for agentic artificial intelligence, from increasing staff productivity to reducing data fragmentation, but providers need a reliable data foundation with strong governance, according to the article. Christian Lindmark, vice president and CTO at Stanford Health Care and the Stanford University School of Medicine, said AI can increase efficiencies but the key is increasing productivity for the organization, which is where governance comes into play. Faila Santos, senior digital health adviser at HIMSS and a family nurse practitioner, said Fast Healthcare Interoperability Resources (FHIR) application programming interfaces (APIs), U.S. Core Data for Interoperability (USCDI) health data classes and data elements, and terminologies such as SNOMED CT and LOINC provide an essential base, and that fragmentation does not disappear but these standards can reduce the need to rebuild the plumbing for every new task. In 2025, Stanford Health Care partnered with Microsoft to use its healthcare agent orchestrator to reduce fragmentation and produce insights from data that was challenging to search, choosing thoracic tumor boards as the first use case. Stanford Health Care uses Microsoft Azure for cloud infrastructure, builds its own AI agents, and Databricks houses its data lake. AI agents can now automate creation of patient summaries before tumor board meetings. Stanford Health Care began with a pilot implementation that ran for about three months, then moved it to full production use after feedback. Anurang Revri, vice president and chief enterprise architect at Stanford Health Care,