The Azure SRE Agent it's more than another AI feature. It's an autonomous system that investigates incidents, builds persistent memory from every interaction, and acts within defined guardrails while you're doing something more useful.
In this session we walk through everything that matters before you deploy your first agent: the cross-subscription architecture pattern that keeps agent infrastructure cleanly separated from managed workloads, the identity model and where the agent actually shows up in Entra ID (spoiler: not where most people look), the knowledge and memory system and how to build a runbook repository your agent will actually use, and what it concretely costs — broken down by model choice.
This is a session based on real deployments. You'll leave with a clear picture of where the Azure SRE Agent delivers today, where it still has limits, and how to get started without surprises.