As Agentforce Agents become mainstream, designing and building them effectively requires more than just creativity — it requires discipline. In this session, we’ll explore proven best practices for designing, building, and managing Agentforce agents that deliver consistent, trustworthy results.

We’ll also share insights on measuring agent performance, optimizing workflows, and aligning AI behavior with your business goals.

You'll learn ten practical practices that separate successful agentic implementations from expensive experiments—from mapping decision rights and instrumenting reasoning paths, to engineering robust fallbacks and building real operational capability. 

We'll explore why most organizations underinvest in resilience engineering, how to test for failure modes systematically, and what it means to govern agentic systems with both speed and responsibility.