How to govern an agentic workforce in production
The guardrails that make it a defensible, compliant solution.
Getting one AI agent to work is easy. Running a whole workforce of them in production, in a way your risk team will sign off on, is the hard part. Autonomy without control is a pilot, not a compliant solution.
Agentic AI earns its place in production only when it's governable. That means a control layer around the agents: people in the loop at the right moments, confidence-driven routing, role-based limits on what each agent may and may not do, clear owners and SLAs for every run, and a full audit trail of what was decided and why. The guardrails are built into the platform, not bolted on afterwards.
This session shows agentic AI in action, the control layer around it and how it holds up as the work scales - and what that means for the AI and operational-resilience rules your risk team will check in every major jurisdiction.
Join Tom Coppock, Sr. Director, AI Strategy & Research, and Jesper Scherpenhuijsen, Principal Solutions Consultant, and learn how to:
- Build an agentic workforce with clear tasks, boundaries and agent-to-agent handoff
- Put the guardrails in place, with human oversight, a full audit trail, clear owners and SLAs
- Know where not to put an agent yet, and why
- Keep it compliant with the AI and operational-resilience regulations emerging worldwide
- See how it all applies in a real TotalAgility use case: loan onboarding
Register now to see how to take agentic AI from pilot to production, defensibly.
