Kube Intelligence - A Metric Based Insightful Remediation Recommender

Explore a conference talk that presents a real-world case study on building an internal platform to address post-deployment challenges in Kubernetes environments. Learn how Google developed a metric-based system to identify and recommend optimizations for applications already in production, addressing the common problem where initial design estimates differ significantly from actual usage patterns. Discover how this intelligent remediation system tackles rightsizing opportunities, identifies candidates for architecture migrations such as moving to serverless platforms, and determines optimal maintenance windows using comprehensive metrics analysis. Understand the methodology behind detecting inefficiencies that often remain hidden until performance degradation becomes apparent or costs become unjustifiable, and see how seemingly minor optimizations can deliver significant impact across system performance and resource utilization.