AI Computer Vision at the Edge

Welcome to the AI Computer Vision Validated Pattern workshop.

This lab guide walks you through a multi-cluster deployment that combines Red Hat OpenShift Container Platform (OCP), Red Hat OpenShift AI (RHOAI), Red Hat Developer Hub (RHDH), and edge inference with NeuroFace computer vision workloads.

What you will explore

  • Hub-spoke fleet management with Red Hat Advanced Cluster Management (RHACM)

  • AI inference at the edge with OpenVINO Model Server (OVMS) and YOLO

  • Cross-cluster connectivity with Skupper and Red Hat Connectivity Link (RHCL)

  • Native Models as a Service in OpenShift AI 3.4 (Gen AI Studio, MaaS gateway)

  • Argo CD MCP for multi-cluster GitOps queries (OpenShift Lightspeed client)

  • Observability with Grafana, Kiali, and OpenTelemetry

  • Developer self-service with Developer Hub, GitLab, and OpenShift DevSpaces

Workshop access

Use your assigned workshop user (user1 through user30) with password Welcome123! to log in to:

These credentials are for this ephemeral sandbox environment only. Do not reuse workshop passwords in production. Sandbox clusters are automatically destroyed after the session.

  • OCP Console (HTPasswd identity provider workshop-users)

  • Red Hat Developer Hub (Keycloak realm backstage)

  • GitLab and DevSpaces (provisioned per user)

If you have not registered yet, open https://workshop-registration.apps.cluster.example.com and enter your email to receive a userN assignment.

Your assigned user (guest (register first)) appears in the badge at the top of this guide after registration. Showroom injects apps.cluster.example.com, %EAST_DOMAIN%, and %WEST_DOMAIN% from the pattern deployment — use them in terminal commands and URLs throughout the lab.

Lab terminal

The embedded terminal on the right supports helper commands:

hub-login guest (register first)
east-login guest (register first)
west-login guest (register first)
neuroface-cv-status guest (register first)
neuroface-cv-traffic
neuroface-spoke-status guest (register first)
maas-native-status
argocd-mcp-status

Next steps

Continue with Lab overview or review the CRD reference.