What Just Happened in Class? Personalized Teaching Feedback from AI

As institutions seek to support effective teaching at scale, faculty often lack timely, personalized feedback about their own classroom practices. This session presents findings from a research initiative that uses generative AI to provide instructors with private, class-session-specific feedback generated from class transcripts, slides, and teaching materials. The system identifies which course learning objectives were addressed in each session and estimates how much time was devoted to them. It also generates sample questions students should be able to answer, highlights which educational theories were employed (with time-stamped links to representative moments in the session), and suggests areas that may have caused student confusion based on prior course content and student level. We share insights from pilot studies exploring how much information instructors find useful, when and how they prefer to receive it, and how the feedback has informed course revision. The goal is to offer a sustainable, scalable mechanism for formative, instructor-centered feedback that improves reflective teaching and course alignment—without requiring additional classroom observation. Attendees will learn how this approach supports teaching development in real time, respects instructor autonomy, and promotes data-informed course design.