Moving from "chatting" with AI to building a functional "agent" requires a shift in mindset from prompt engineering to systems design. This session shares a case study of developing a "course architect" agent capable of transforming course materials into a comprehensive 16-week instructional ecosystem. By grounding the agent’s outputs in specific instructor content rather than the generalized training of general-purpose LLMs, the architecture ensures pedagogical authority of the instructor. The session will demonstrate a workflow that generates a robust suite of LMS-ready outputs, all validated against national quality assurance benchmarks. The session addresses the "developer’s dilemma" by offering a comparative evaluation of leading GenAI platforms in developing purpose-built “course-architect” agents.
