Going Beyond Infrastructure Observability: Meta’s Approach

In a discussion on the OpenObservability Talks podcast, Dotan Horovits and David Ostrovsky explore the transformative role of observability at Meta, particularly how it shifts from infrastructure-focused tasks to business-driven objectives aimed at enhancing both external and internal user experiences. Ostrovsky, a software engineer at Meta, contrasts his experiences at Meta with those at his previous company, Proofpoint, noting that while observability at Proofpoint was primarily engineering-driven, at Meta it starts with business discussions around service level agreements (SLAs) and focuses on the decision-making layer in fraud detection. Meta’s approach allows teams to leverage a robust infrastructure managed by separate dedicated teams, enabling Ostrovsky’s team to concentrate on higher-order observability that directly impacts product performance and user experience. This approach underscores the importance of addressing business-facing decisions through intelligent data analytics over telemetry data, which is key to enhancing developer observability and accelerating development cycles.