Kubeflow: Cloud-Native Machine Learning on Kubernetes

Join us for a talk and demo on the Kubeflow project given by Jeff Fogarty of U.S. Bank. Kubeflow is a machine learning platform that runs atop Kubernetes, based on Google’s internal machine learning pipelines.

https://www.kubeflow.org
https://github.com/kubeflow/kubeflow

Jeff's talk will be followed by a demo, by the event sponsor Arrikto, on the minikf project, which installs Kubeflow locally on the workstation.

https://www.arrikto.com/minikf/

Arrikto is a CNCF member and makes significant contributions to cloud-native storage options on Kubernetes.

https://www.arrikto.com

Abstract:

With the popularization of Data Science across all industries, the ability to have an efficient, composable, portable and scalable production Data Science environment is being demanded at the highest levels of the enterprise.

In this talk, we will introduce Kubeflow, the leading containerized machine learning platform designed to leverage efficient Kubernetes operations. We will show how to run Kubeflow on a laptop, an on-prem infrastructure and on Google Cloud.

We will also give a high-level overview of features within Kubeflow including Jupyter Notebooks, Katib, and Pipelines.

Speakers:
Jeff Fogarty, Technology Architect @U.S. Bank