How Subaru Accelerated AI Model Development for Next-Generation EyeSight...- Ryoji Kobayashi, Subaru

This talk explains how Subaru built a Kubernetes-based AI model development platform to accelerate next-generation EyeSight driver-assistance AI development. The team addressed challenges with large ML container images (reduced pull time from 3 hours to 3 minutes), manual deployment operations (GitOps for 25 applications), and complex ML workflows (automated with Argo Workflows). Target audience includes DevOps engineers and ML platform teams building production ML infrastructure.