Autonomous Feedback Loops for Performance and Reliability - SRE Track

Most engineering teams address performance and reliability issues as one-off incidents, so their use of AI follows the same pattern: adhoc requests to investigate or fix problems as they’re encountered. But what if it could be more autonomous—a systematic process leading to continuous improvement?

In this fireside chat, Martin Amps, member of technical staff at Anthropic, and Alan Gates, Senior Staff Engineer at Datadog, compare notes on how they're putting AI agents into continuous feedback loops to drive real operational gains. They’ll share how autonomous agent loops can mine telemetry for optimization opportunities, explore potential action plans based on the data, and even improve the telemetry itself.

You’ll leave with a clear plan for setting up your own agent-driven feedback loops to continuously improve system performance and reliability.