AI demos typically showcase an ideal workflow. Healthcare leaders must evaluate what happens beyond that controlled environment: how a solution handles real-world data, edge cases, poor outputs and workflow variation. In this sponsored episode of Podnosis, Chad Dodd, Vice President of Product Management at athenahealth, discusses the evidence clinicians and practices need before trusting an AI-assisted tool.
Dodd explains why governance should extend across the product lifecycle, including testing for bias, accuracy and data validity; validation across specialties, datasets and patient samples; and clear identification of AI-generated content. He also outlines how links to source information can make clinician oversight more efficient without shifting the burden of full-time validation to physicians.
The discussion also addresses post-deployment monitoring, including evaluation pipelines, model drift, telemetry, feature usage and user feedback. For organizations assessing AI partners, Dodd recommends focusing on demonstrated workflow value, data privacy and security, transparency around failure points, and evidence that the technology performs safely and compliantly. Listen to the full podcast to learn more.