Skip to main content

What is Kubeflow?

Kubeflow is an open-source platform designed to make deployments of machine learning (ML) workflows on Kubernetes simple, portable, and scalable. It provides a comprehensive stack of tools and services to support the entire ML lifecycle, from interactive development to model training, serving, and monitoring—all within a Kubernetes-native environment.

Kubeflow brings together best-in-class components for data scientists, ML engineers, and DevOps teams, enabling them to:

  • Develop and run Jupyter notebooks in a secure, scalable environment
  • Build, manage, and execute reproducible ML pipelines
  • Serve and monitor machine learning models at scale
  • Integrate with popular IDEs like VS Code for cloud development
  • Choose from different Kubeflow flavours to suit specific needs

What does the Kubeflow stack offer?​

Kubeflow provides a modular set of features, each accessible through dedicated interfaces and APIs:

  • Notebooks: Launch and manage Jupyter notebooks for interactive data science and ML development.
  • Pipelines: Author, deploy, and manage end-to-end ML workflows using Kubeflow Pipelines.
  • KServe: Deploy, serve, and monitor ML models with advanced inference capabilities.
  • VSCode CDE: Use Visual Studio Code in the cloud for collaborative development and experimentation.
  • Kubeflow Flavours: Explore different Kubeflow distributions and deployment options.

Kubeflow is designed to be flexible and extensible, supporting a wide range of ML frameworks and tools, and integrating seamlessly with Kubernetes infrastructure.

Explore the sections above to learn more about each core feature of Kubeflow.