Main Conference Agenda 2025

Running workloads on the edge means operating with a finite amount of compute resources, which means that running workloads as lightweight as possible is generally preferable. In contrast, securely operating workloads in the cloud often means spinning up additional virtual machines for strong workload isolation. On the edge, this strategy is often not feasible. This talk covers how you can securely run containers in lightweight Kubernetes clusters on the edge entirely without the need for virtualization. This talk will cover different hardening approaches and compare them with special focus on securing AI workloads on the edge.