Build a machine learning stack on Kubernetes using Kubeflow

Description

Kubernetes has quickly become the hybrid solution for deploying complicated workloads anywhere. Although Kubernetes started with just stateless services, customers have begun to move complex workloads, such as those for machine learning, to the platform, taking advantage of its rich APIs, reliability, and performance. However, building any production-ready machine learning system involves various components, often mixing vendors and hand-rolled solutions. Connecting and managing these services for even moderately sophisticated setups introduces huge barriers of complexity in adopting machine learning. Infrastructure engineers will often spend a significant amount of time manually tweaking deployments and hand-rolling solutions before a single model can be tested. Worse, these deployments are so tied to the clusters they have been deployed to that these stacks are immobile, meaning that moving a model from a laptop to a highly scalable cloud cluster is effectively impossible without significant rearchitecture.

Nilesh Patel explains how to use Kubernetes as a platform to run machine learning apps, using Kubeflow, a new open source project launched by Google dedicated to making using ML stacks on Kubernetes easy, fast, and extensible.