Speakers
Mihai Criveti, Distinguished Engineer - Agentic AI, IBM
AI agents are moving from assistants to autonomous systems that can access data, invoke tools, call APIs and collaborate with other agents. That creates a new enterprise challenge: knowing what agents exist, what they can do, who they act as, how they behave, what they cost, and whether they remain safe, compliant and valuable.
This talk explores the Agent Control Plane: a governance and operations layer for discovering, securing, observing and operating agents across models, frameworks and clouds. Drawing on watsonx Orchestrate and the open-source ContextForge project, Mihai will show why identity, policy, observability, AgentOps, lifecycle management, FinOps, risk controls and open standards such as MCP and A2A are becoming essential infrastructure for enterprise AI - helping organizations address requirements such as the EU AI Act while proving ROI, controlling cost and scaling agent adoption responsibly.
