PathShaper Launches for BADM 554

AgentLab deploys PathShaper, an adaptive learning platform powered by multi-agent content generation, for BADM 554 at Gies College of Business.

PathShaper, the production pilot of our Learning 3.0 architecture, is now live for BADM 554 (Enterprise Database Management) at Gies College of Business. The platform introduces a new model for course content creation where faculty define intent and AI agent swarms generate personalized learning materials.

How PathShaper Works

PathShaper follows a four-step workflow that keeps faculty in control while leveraging AI for content generation at scale:

  • NLSpec Definition — Faculty write natural language specifications describing learning outcomes, concepts, and pedagogical intent
  • Agent Generation — Coordinated multi-agent swarms generate textbooks, quizzes, labs, and studio guides from the specs
  • Faculty Review — All generated content goes through faculty approval before students see it
  • Student Personalization — Students navigate a 186-concept learning graph with adaptive content tailored to their pace

Faculty as Shapers: PathShaper redefines the faculty role from content creator to "shaper on the path" — defining the learning journey while AI handles the heavy lifting of content production.

BADM 554 Pilot

The initial deployment covers 8 weeks of enterprise database curriculum with 186 concepts, 39 content items, and 9 learning outcomes. Students interact with an interactive knowledge graph that visualizes their learning progress and adapts to their individual needs.

Technology Stack

PathShaper is built with Next.js 15, React 19, and TypeScript, with vis-network powering the interactive learning graph visualization. The platform is deployed on Vercel for fast, reliable access.

Visit the PathShaper project page to learn more about the architecture and roadmap.