Skip to Main

Health Technologies:

LLM-Powered OSCE Grading: The Future of Clinical Skills Evaluation

Andrew Jamieson, Ph.D.

  • Assistant Professor, Lyda Hill Department of Bioinformatics, UT Southwestern Medical Center

The Jamieson Lab

Every year, thousands of medical students take clinical exaM.S. that are time-consuming to grade, inconsistent across evaluators, and offer limited feedback. Dr. Andrew Jamieson and his team have built a game-changing solution: a secure, zero-shot AI system that uses large language models to automatically assess OSCE (Objective Structured Clinical Examination) performance, including written notes, audio, and video. The platform matches or exceeds human graders in accuracy, reduces turnaround time from weeks to days, and cuts evaluator effort by over 90%. Already validated and deployed at UTSW, the system is gaining attention from medical schools across the country eager to modernize clinical education. With institutional backing and a commitment from UTSW to serve as its first paying customer, this venture is now poised for commercial launch.

Stage 3: Testing and Validation