mlops-for-devops

Why we included this project

Most MLOps material assumes you are a data scientist learning infrastructure. This repo flips that: it is written for DevOps, platform, and SRE engineers who want to run ML workloads without becoming data scientists. You follow one employee-attrition prediction problem from local data pipelines and model training up to enterprise orchestration with Airflow, DVC, Kubeflow, MLflow, KServe, and Evidently on Kubernetes. The phases build on each other, the code is in the repo so you can run each step, and the explanations lean on DevOps analogies rather than ML theory. If you already know Kubernetes and Docker but have not touched MLOps, this is a practical bridge into the tooling you will actually meet in production.