Ashley Gross

The interesting question about an enterprise agent is not what it can do. It is what it is allowed to do when nobody is watching, and who answers for it afterwards.

AI agents change how work gets done, but only when they are designed around real workflows, clear guardrails, and accountable owners. This session moves past chatbot experimentation into operational design: which processes suit agent-assisted work, how to define roles and boundaries, where approval gates belong, and how to improve speed without giving up trust or accountability.

Best for AI, automation, operations, IT, business process, innovation and technology audiences.

Three things attendees will learn:

1. How to identify which business processes are genuinely suitable for agent-assisted work, and which only look suitable.
2. How to define agent roles, boundaries, approval gates and measurement before deployment rather than after an incident.
3. The over-automation mistakes that force a rollback six months in, and how they get scoped in on day one.