New Capabilities to Observe Agents in Azure Monitor (opens in new tab)

Over the last six months, we have been listening to you and building new capabilities to help you observe your agents. You’ve been sharing with us that quality issues are tricky and evaluation is critical, that agent reasoning needs to be understood, that humans must be in the loop to review select agent interactions, and that security and privacy are essential. To address these concerns, we’re announcing several new capabilities that make agents a first-class artifact in Azure Monitor, so yo...