This document discusses Kubernetes Event-driven Autoscaling (KEDA), which allows applications running on Kubernetes to automatically scale based on external events. KEDA manages workloads to provide autoscaling to zero, registers as a custom metrics adapter, and provides metrics for the Horizontal Pod Autoscaler to use for scaling. It supports a variety of event sources and scalers out of the box. KEDA is cloud-agnostic, vendor-neutral, and easy to install via Helm charts or the Operator Framework. It has over 1,800 stars on GitHub and an active community. The document demonstrates how KEDA works and discusses its roadmap and integration with technologies like Knative and Azure Functions.