Learn about building eBPF-powered observability solutions for sustainable computing in this conference talk from Google engineers. Explore how organizations deploying AI/ML workloads on Kubernetes can address environmental impacts from energy-intensive model training. Discover open-source cloud agnostic tools and techniques for monitoring performance and energy metrics from Kubernetes resources like Pods and Nodes. Master methods for optimizing AI/ML workloads to reduce energy consumption, lower operational costs, and minimize carbon emissions while maintaining high performance standards. Gain practical insights into implementing FinOps practices and sustainable computing approaches as AI/ML deployments continue to scale across cloud environments.