This talk dives into the technical architecture and operational strategies behind Elastic's global-scale SLO management system, designed to handle hundreds of thousands of SLOs across 60+ regions and major cloud providers. The system empowers development teams to define, monitor, and manage SLOs autonomously, fostering accountability and operational excellence.

Key topics include the phased rollout of Elastic’s Serverless product and the foundational role of an "Observability as a Service" platform. This platform integrates SLO burn rate checks, dashboards, alerting mechanisms, and workflow execution systems to streamline operations and enable rapid troubleshooting. The talk also explores how these technical solutions, combined with revamped incident management processes, ensure service reliability while reducing SRE involvement.

Learn how Elastic re-architected its production readiness by co-developing quality gates with engineering teams, integrating SLO checks in software delivery, and leveraging analytics to enable data-driven operational reviews. These efforts are complemented by training development teams to take full ownership of on-call duties and incident management, ensuring the system’s scalability and resilience.

Attendees will gain insights into the design principles, challenges, and best practices behind building a self-service SLO management system that supports global SaaS operations.

Panagiotis Moustafellos is a systems engineer with over 20 years of experience in diverse tech environments. His areas of expertise include systems architecture, observability, and security, as well as scaling software systems and infrastructure. Currently he is a Distinguished Engineer at Elastic, building observability products and driving the production readiness transformation for Elastic Cloud.