The growing number of applicants to medical schools is creating a significant workload for admissions committee (AC) members, particularly when a holistic review is used, which includes a comprehensive assessment of academic metrics, experience, attributes, and competencies. To address this, we have operationalized artificial intelligence (AI) and machine learning to support AC members throughout the applicant review process. For some medical schools, serving on the AC is an uncompensated role, accounting for hundreds of hours annually. AI and machine learning can offer opportunities to increase efficiency in the applicant review process while maintaining the spirit of comprehensive, holistic reviews, provided they are created and implemented appropriately (Phase 1).

AI can also be used to create executive summaries for interviewed applicants, allowing AC members to access immediate data and inform admission decisions more effectively (Phase 2). Additionally, deploying a chatbot that enables AC members to elicit additional data beyond that provided in the executive summary may be beneficial during the evaluation process.

Using a phased approach to AI product deployment in medical school admissions has resulted in approximately a 150-hour decrease in workload annually for AC members. We anticipate reducing 75 additional uncompensated hours for AC members during Phase 2. Our outcomes may inform AI-driven admissions-related initiatives at other medical schools.

In this session, we will share our experiences in developing and deploying a machine learning screening tool to review medical school applicants, as well as additional admissions projects leveraging artificial intelligence that are currently in development. This session is relevant because it addresses assessment, and the data collected during the admission process serves as the initial information used to create precision education tools that support students in their medical education journey.