Sep 5, 6:00βββ7:30β―AM (UTC)
Event time displays in your timezone.
2055 RSVPs
Session 5 brings it all together. We've built, connected, orchestrated, and prepared our agents for production β now it's time to govern them. We'll explore the controls, permissions, human oversight, traceability, and guardrails enterprises need to deploy agentic AI with confidence. Because powerful agents aren't enough they need to be trusted. π€
1. From Agentic AI to Governed Agentic AI
Reconnect the 5-session journey and understand why governance isn't something you add after deployment β it needs to be designed into the architecture from the beginning.
2. The Agent Governance Framework
Define the policies, controls, ownership models, and guardrails required to govern autonomous and semi-autonomous agents across the enterprise.
3. Identity, Access & Agent Permissions
Understand agent identities, authentication, least-privilege access, tool permissions, secrets management, and how to control exactly what an agent can see, decide, and do.
4. Human-in-the-Loop & Approval Controls
Design approval layers, escalation paths, confidence thresholds, exception handling, and human checkpoints for high-impact agent decisions and actions.
5. Observability, Auditability & Traceability
Build the audit trail. Understand how to capture prompts, model responses, tool calls, agent decisions, actions, approvals, versions, and outcomes so you can answer: What happened, why did it happen, and who was responsible?
6. AI Trust Layer
Explore the controls that sit between agents and enterprise systems β policy enforcement, validation, security, monitoring, model and prompt controls, data protection, and runtime guardrails.
7. Risk, Compliance & Responsible AI
Identify agent-specific risks including hallucinations, excessive agency, data leakage, unauthorized actions, model drift, prompt injection, and unintended business outcomes β and understand how governance reduces the blast radius when something goes wrong.
8. Operating Governance in Production
Move governance from a PowerPoint into an operating model: ownership, monitoring, incident response, rollback and kill-switch mechanisms, change management, periodic evaluations, and continuous improvement.
9. Your Role in the Agentic Enterprise
Connect your RPA and automation experience to emerging roles across agentic architecture, AI governance, orchestration, evaluation, and intelligent automation leadership β and identify the skills you should build next.
10. Final Program Recap & Your Next Move
Bring all five sessions together β from foundations and tooling through orchestration, production, and governance β and turn what you've learned into a practical roadmap for your first enterprise agentic AI initiative.See you in Session 4! π
NatWest Group
Vice President - RPA & Analytics
Iris software
Solutions Consultant
RPA Manager at Boundaryless Group
UiPath MVP
UiPath
Community Lead - Asia Pacific & Japan