In this episode, we explore how artificial intelligence and machine learning are advancing the use of electronic health record data in rheumatology. Dr. Kat discusses the development and validation of a natural language processing–based model to infer rheumatoid arthritis disease activity from structured and unstructured clinical data, its application in studying cardiovascular outcomes, and the challenges of deploying predictive models across different health systems. The conversation also examines the future of real-world evidence generation, the practical limitations of AI in clinical research, and Dr. Kat's career journey, offering insights and advice for the next generation of physician-scientists.