Graduate students in AI courses face a paradox: the very technology they are studying is generating an overwhelming flood of learning materials, making it harder, not easier, to learn. For working professionals balancing demanding careers, families, and graduate coursework, "overwhelm fatigue" is real — and it is undermining engagement with one of the most important topics of our time. This session introduces a structured solution: a Claude-powered learning repository designed to give time-constrained graduate students a navigable, personalized entry point into generative AI concepts. Rather than confronting students with an undifferentiated mass of videos, blogposts, and references, the repository uses carefully designed prompts to help each student find their own path through the material. The session includes a live demonstration — attendees will watch an actual interaction between the presenter and Claude in real time — and participants are invited to engage directly during the session. Attendees will leave with a replicable framework, available as a downloadable PDF, that faculty across disciplines can adapt to build their own Claude-powered course repositories.