A Year of Democratizing ML With Kubernetes & Kubeflow

Explore the evolution of the Kubeflow project in this 36-minute conference talk by David Aronchick and Fei Xue from Google. Discover how Kubeflow has become the most popular ML platform for Kubernetes, enabling composable, portable, and scalable end-to-end ML pipelines. Learn about the newly announced Kubeflow 1.0 and its features, including a web-based UI, simplified setup, and sophisticated ML tooling such as hyperparameter search and Google's TensorFlow Extended project. Gain insights into the integrated Pipelines project, which enables multi-cloud ML with continuous training and hosted services. Understand how Kubernetes native extensibility is making ML accessible to a broader audience, allowing anyone who can code to build complete end-to-end solutions.