Faculty spend hours writing the same feedback on student work often with little impact on learning. This talk demonstrates how to design a course-specific AI feedback agent that delivers consistent, rubric-aligned, pedagogically grounded feedback while preserving academic judgment. The session shows how teaching intent, assessment criteria, and tone are translated into agent behavior in real time, without coding. Participants see a live walkthrough of the agent’s structure and leave with a reusable pattern they can adapt for writing, projects, problem-solving, or reflective assignments across disciplines.
