SLOs have two hard moments: deciding what to measure when a service is new, and working out why an error budget is burning once it is live. This session is about two things AI can now do with those moments. Both run live on real systems in Nobl9, with time for your questions.
The first is turning a conversation into an SLO. You describe a service in plain language, and an AI interviews you about its user journeys and failure modes the way an experienced SRE would. It comes back with a drafted SLO: a recommended indicator, a target, and the reasoning behind the number. You review it and adjust. The blank page is gone.
The second is asking why a budget is burning and getting a real answer. When an alert fires, AI investigates your own traces and metrics, finds the request, user, or response behind the burn, and writes the explanation next to the alert. Sometimes the answer is that the alert is just noise, which is worth knowing too.
Bring the SLO question your team keeps arguing about. Brian Singer and Andrzej Voss will work through it live.
