MLOps Pipelines: Kubeflow & Argo

About This Course

A hands-on module for infrastructure engineers moving from DevOps into MLOps. Starting from what you already know — CI/CD, containers, Kubernetes — it builds the ML delta step by step: why ML breaks classic CI/CD, the anatomy of training pipelines as DAGs, Argo Workflows as the Kubernetes-native pipeline engine (templates, artifacts, retries, CronWorkflows, Argo Events, GPU steps), Kubeflow Pipelines for Python-authored pipelines with lineage and caching, and GitOps delivery of models with Argo CD — repo layout, registry-to-production promotion, progressive delivery with Argo Rollouts, and the drift-triggered continuous-training loop, closed out with pipeline observability.

Each of the four modules ends with a graded knowledge check of scenario-based questions. Best suited for DevOps/platform engineers who have completed the NCP-AIO study track or have equivalent Kubernetes and GPU-cluster experience.