#235 - Doctors + HR Embrace Smart AI

Inside Advocate Health's AI Strategy — From a Physician Who Builds It

Featuring Dr. Andy McWilliams, Clinical AI Product Development Lead, Advocate Health

Recorded live at Advocate Health's The Pearl innovation quarter in Charlotte, North Carolina.

Dr. Andy McWilliams is a practicing hospitalist — boarded in internal medicine and pediatrics — who has spent 15 years at Advocate Health and now helps lead its clinical AI strategy. He is not a theorist. He is a physician who uses these tools, sees patients, and builds governance around the same technology his colleagues depend on every day.

🏛️ Governance First — Before the Hype Advocate's AI governance framework came before the widespread deployments, not after. The framework was built collaboratively across ethics, legal, compliance, data science, and HR — with a shared understanding of the why before the how. Every AI solution entering the organization, regardless of department, follows a tiered risk review. HR evaluating a new HRIS with embedded AI? It goes through the same pathway as a clinical solution.

🎙️ Ambient Scribes: 6,000 Clinicians, Measurable Results Advocate was among the first health systems to partner with Microsoft on ambient AI scribing. Today roughly 6,000 physicians and APPs use it regularly. Copilot Dragon for nursing is in early release. The outcome is not just efficiency — it is eye contact, presence, and a patient experience that changes because the clinician is no longer turned toward a keyboard.

📷 Room of the Future: Cameras, Microphones & Virtual Nursing Advocate is installing cameras and microphones across inpatient facilities to enable virtual nursing, hands-free documentation, real-time fall risk alerts, and seamless family connection. Patients report feeling an additional layer of security — not surveillance — when the system is explained clearly.

⚠️ The Heebie Jeebies Are Real — and Healthy Dr. McWilliams is direct: black-box AI in clinical decision-making without explainability gives him pause. The democratization of health information via LLMs has clear benefits, but solutions deployed without guardrails in healthcare concern him. His three questions for any organization evaluating AI: understand the use case and its risks, know the worst-case scenario, and audit for potential bias.

Connect with Dr. McWilliams: https://www.linkedin.com/in/andy-mcwilliams-52254a5