Before You Deploy AI: Mapping Healthcare Workflows to Find What Should (and Shouldn't) Be Automated
Most healthcare AI discussions focus on models, accuracy, and compliance. Far fewer address the human decisions, trust signals, and workflow changes that determine whether AI succeeds in practice. This session introduces The Human Middle framework, demonstrating how healthcare organizations can design AI-enabled workflows that balance automation with human judgment, reduce friction, improve adoption, and strengthen governance without slowing innovation.
About the speaker
Joan Kelly
Founder and CEO at Caidir
Joan Kelly, EdD, MBA, is a healthcare executive, experience architect, AI strategist, and founder of Caidir, an AI decision platform helping healthcare organizations design, govern, measure, and improve AI-enabled care delivery. Over the past two decades, Joan has led consumer, patient, and workforce transformation initiatives for some of the nation’s most recognized healthcare organizations, including Yale New Haven Health, NYU Langone Health, Humana, and Press Ganey. Her work has focused on redesigning care journeys, improving outcomes, advancing patient and staff experience, and translating innovation into measurable operational impact. Today, Joan advises health systems, healthcare technology companies, and executive teams on AI strategy, workflow transformation, governance, and human-centered implementation. She is the creator of The Human Middle framework and author of the forthcoming book Orchestrated Experience, which explores how organizations can deploy AI responsibly while strengthening trust, performance, and human connection. Joan is recognized for her practical approach to healthcare AI, bridging strategy, operations, experience design, and governance to help organizations move from AI experimentation to sustainable transformation.