AI inference at the edge using OpenShift AI

Many organizations are looking to edge deployments of AI to provide real-time insights into their data. OpenShift AI can be used to deploy model serving, predict failure, detect anomalies, and do quality inspection in low-latency environments in near real-time. The demo shows how the model can be packaged into an inference container image and use data science pipelines to fetch models, build, test, deploy and update within a GitOps workflow. ArgoCD is used to detect changes and update the image at the edge devices if needed. Observability into the model’s health and performance is provided by gathering metrics from edge devices and reporting back to centralized management.