12:30 - 1:00 PM (ILCB 207)
Presented by: Vincent VanBuren (Naresh K. Vashisht College of Medicine)
As institutions redesign curricula for flexibility, flipped classrooms, and competency-based progression, lecture capture remains anchored to an outdated production model: record, edit, upload, repeat. This approach is labor-intensive, difficult to update, and poorly aligned with modern digital learning design. This project replaces recording with structured content engineering. We developed an AI-enabled pipeline that transforms PowerPoint slide decks into modular instructional videos using cloned faculty voice, tagged slide segmentation, automated narration synthesis, caption generation, and video rendering. Rather than capturing live delivery, the system parses presenter notes and embedded delimiters to generate reusable, precisely segmented modules from a single source deck.
The workflow was deployed in production across the complete Evidence Based Medicine module series within a medical curriculum. All module videos were generated using this pipeline. Students praised the clarity and pacing of the videos, indicating that authentic instructor presence was preserved while production burden was dramatically reduced.
This design reframes AI not as a content author, but as instructional infrastructure. Faculty maintain intellectual ownership and pedagogical intent while gaining the ability to rapidly update, modularize, and scale content across courses. Captioning support and structured segmentation enhance accessibility and reuse without additional manual effort.
By shifting from lecture capture to learning architecture, this model supports curricular agility, faculty sustainability, and long-term digital readiness. It offers a scalable framework for designing learning environments that meet the demands of today while preparing institutions for tomorrow.