Build and Manage Kubernetes-based Machine Learning environments

Strategic executive decisions are migrating from intuition-based decision making to data-driven decision making. The adoption, of Open Source Machine Learning (ML) solutions, is continuing to grow, and ML teams are striving to quickly deliver (and justify) reliable solutions with modern functionality.

This presentation will review the critical questions that Fortune 500 customers are asking Canonical regarding how to effectively build and manage Kubernetes-based ML environments. The talk will review Kubernetes Market Momentum, Business Drivers, Use Cases and a Definition of Success for a K8s deployment. It will review the Core Operations, Portability and Interoperability requirements for a Kubernetes ML stack and touch on new R&D efforts, including simplifying Machine Learning (with Kubeflow). It will also review common technical concerns, as well as vendor support & exit options. A live demonstration of an ML environment with JupyterHub, Tensorflow, Kubernetes and Ubuntu will be provided.
Josh Bottum is a Canonical Regional Manager for Large Enterprise customers. Canonical supports market leaders during their journey to design, implement and justify Open Source based solutions. During the last 30 years, Mr. Bottum has held sales, business development, and product management positions at Canonical, Cisco, NetApp, Ascend, Lucent, AT&T and a handful of successful and failed start-ups.