Welcome to the AI Agent & Copilot Podcast, analyzing the latest AI Copilot and agent developments from Microsoft and its partners, delving into customer use cases, and exploring how AI plus the Cloud helps customers reimagine their business. In this episode, Tom Smith speaks with Michael Pencina, Ph.D., Chief Data Scientist at Duke Health, and Dave Galich, Advisory Practice Lead and Chief Product Officer for Smart AI Governance Engine (SAIGE) at Avanade. Duke Health is using SAIGE and collaborating with Avanade in its development.
Highlights
Duke Health Background, AI Use Cases (00:46)
Duke Health is a premier academic medical center, part of Duke University, with state-of-the-art care delivery. Pencina highlights Duke Health’s leadership in AI adoption, citing a strategic partnership with Microsoft. Duke Health is also an early adopter and development partner for ambient voice recognition technology from Abridge.
Microsoft Infrastructure, AI Implementation (03:28)
Pencina details Duke Health’s use of Azure cloud services, Microsoft Fabric for data management and analytics, and Microsoft Azure AI Studio. Duke Health is piloting Microsoft’s multi-modal AI algorithms for data integration from various sources.
Avanade AI Governance (04:46)
Galich noted Duke Health reached a plateau in developing its own AI governance, managing it with Excel spreadsheets and other tools. Pencina discusses early days of AI adoption at Duke Health, the chaos, and the need for a principled approach to AI governance. Its internally developed governance model was ready for the GenAI revolution, allowing quick adaptation.
AI Agent & Copilot Summit is an AI-first event to define opportunities, impact, and outcomes with Microsoft Copilot and agents. Building on its 2025 success, the 2026 event takes place March 17-19 in San Diego. Get more details.
Components of AI Governance (10:05)
Pencina outlines the three components of AI governance: inventory management, risk assessment and monitoring, and the governance process itself. The inventory component is critical for managing AI algorithms. Risk assessment and monitoring involves risk stratification and continuous monitoring of AI algorithms due to their constant learning and adaptation. The governance process includes identifying clinical and technology owners, accountability, and expertise for deploying AI algorithms.
Development of SAIGE (15:17)
SAIGE aims to make AI governance scalable across various healthcare institutions, including smaller nonprofits. The platform is vendor-agnostic, allowing healthcare systems to choose the best AI tools for their needs.
Current, Future Use of SAIGE (18:52)
Duke Health is in the late stages of piloting the SAIGE platform and is impressed with its value. The platform is expected to be fully operational by year-end, with all AI algorithms integrated into the system. Galich says the platform will include new features and functionality by year-end, enhancing collaboration and information exchange between health systems and AI vendors.
Ask Cloud Wars AI Agent about this analysis
