AI Enabler for PMs Session 2: Practical AI-Orchestrated Delivery

Important note: Register for the event using the Zoom link.

Hey friends, here's the 2nd session of becoming an AI-enabled PM.

Session 1 covered how project and program managers can move beyond basic chatbot usage by applying structured prompting, context engineering, guardrails, harness design, and human-in-the-loop controls.

Session 2 will be a practical, hands-on session focused on applying these concepts to a real program-management workflow.
Using the CloudFirst digital banking migration case study, participants will see how an AI agent can analyse program information, identify risks and dependencies, prepare a program-health summary, recommend actions, and draft stakeholder communication through a controlled multi-step workflow.
The session will demonstrate how to:

• Provide the agent with trusted program context
• Define its role, objective, scope, and operating boundaries
• Design a Plan → Act → Check → Decide loop
• Prevent unsupported assumptions and invented facts
• Set approval gates before any output is published or acted upon
• Review, validate, and improve the agent’s output
• Convert a one-time prompt into a repeatable PM workflow

The objective is not to automate the project or program manager. It is to show how PMs can orchestrate AI to reduce manual coordination work, surface delivery signals earlier, and spend more time on judgement, decisions, stakeholder alignment, and intervention.

### Session Outcome:

By the end of the session, you should understand how to design and run a controlled AI-assisted workflow for risk, dependency, status, and stakeholder-management activities.