Filip Bogdan Serban-Dragan, a second-year Ph.D. student in the Division of Prevention Science and Community Health, delivered a TED-style presentation at the University of Miami’s Eighth Annual Graduate and Postdoctoral Research Symposium on the use of AI to support experiential learning. His work is supported by the Academy of Medical Education Scholars (AMES) and informed by training through the University’s Platform for Excellence in Teaching and Learning (PETAL).

At the center of Serban-Dragan's work is a fundamental question: how can education better prepare students to apply what they learn in real-world settings?

He noted that only a portion of classroom learning is translated into practice, highlighting a persistent gap in how students are trained for applied public health work. His research explores how AI can help address this challenge by creating environments where students actively engage with complex, real-world scenarios.

Reframing Learning Through Simulation

Serban-Dragan's approach uses AI-driven simulations to place students in realistic decision-making environments. Rather than passively learning concepts, students interact with AI-generated personas, such as community leaders, school administrators or health system stakeholders, who present real-world challenges.

Through these interactions, students identify risk and protective factors, propose interventions and receive immediate feedback on their decisions.

“What distinguishes this approach is that it flips the traditional model of education,” he said. “Instead of starting with theory, students begin with the situation, just as they would in real-world public health practice.”

By engaging in these simulations, students develop practical skills, strengthen their understanding of prevention strategies and build confidence in applying their knowledge.

Addressing Gaps in Public Health Training

The work targets two key challenges in public health education: the gap between learning and application and the limited opportunities for students to practice decision-making in realistic settings.

AI-supported simulations create a space where students can test ideas, make decisions and refine their thinking without real-world consequences. This approach allows for repeated practice and insightful learning, which are critical for developing applied skills.

Expanding the Reach of Experiential Learning

Audience discussion following the presentation reflected interest in extending this model beyond higher education. Participants raised questions about how similar approaches could be applied in K–12 settings and other training environments.

The conversation also emphasized the importance of thoughtful instructional design, with attention to how AI-generated experiences can support meaningful learning rather than simply introducing new technology.

Looking Ahead

Serban-Dragan emphasized that AI is not a replacement for instructors or traditional learning environments but a tool to enhance them.

“AI gives us the opportunity to extend the classroom into the real world,” he said. “If we design these experiences well, we can create learning environments that are immersive, practical and this can become one of the most powerful strategies for prevention.”

Written by Deycha Torres Hernández, published March 30, 2026.