A Build Lab Is a Better Test of Learning—If It Ends in Follow-Through

On October 22, the Project Management Institute announced new learning offerings, including an agentic-AI course with a hands-on build lab. PMI says learners will build an agent, diagnose its performance, refine its scope and oversight, and leave with a working agent and a stakeholder brief. (PMI’s announcement)

I like the move from learning about a tool to making something and explaining how it should be governed. But a finished build is a starting signal, not proof that the skill will stick. A useful follow-up would be to revisit the project after a short interval: can the learner still explain the limits, spot a failure, and decide what to change?

What would convince you that a project-based course produced lasting capability—not just a convincing demo?