Machine Learning on Kubernetes at Shell: A Kubeflow Journey

November 18, 2020

November 18, 2020 @ KubeCon + CloudNativeCon North America Boston 2020

In this session, Shell describes the lessons learned from:

  • Working with multiple Machine Learning platforms and tools
  • The challenges of different systems, why we chose Kubeflow
  • How we are now delivering successful models faster and at scale.

Follow our journey as we learned how to deploy highly available, scalable, and secure Kubeflow clusters in the public cloud.

We will describe the lessons learned and steps taken to improve our deployments including enterprise authentication and authorization, network integration, and data science workflows.

We also discuss why we moved away from other platforms and chose Kubeflow, and how it has increased our Data Scientists’ productivity and reduced DevOps overhead.

Today our teams are more self-sufficient, and iterate faster to produce production-ready models in a timely fashion. A zero to hero story made possible by Kubeflow and Kubernetes

Alex Iankoulski

Alex Iankoulski

Technical Leader, ML Orchestration @ Shell New Energies

Alex Iankoulski is a full stack software and infrastructure architect who likes to stay hands-on. He is a Docker Captain who has been helping accelerate the pace of innovation by applying container technologies to solve engineering, data science, AI and ML problems. Alex is currently Technical Leader, ML Orchestration at Shell New Energies where he supports Shell’s ambition to reduce the global carbon footprint. He works to bridge the gap between prototype and product using K8s and Kubeflow. His prior experiences include a tenure at GE Global Research, BHGE, as well as roles in the Healthcare, Network Security and Telecom industries. Alex has presented at conferences like DockerCon, Microsoft Azure AI Conf, Google Next, and is very excited to be a first-time speaker at KubeCon 2020.

Vangelis Koukis

Vangelis Koukis

CTO

Vangelis Koukis is the CTO and founder of Arrikto. Vangelis holds a PhD in computer science and has a long history of working in storage, data management, and cloud computing. Arrikto is a core contributor to the Kubeflow project, mainly in the areas of data management and UX. Arrikto creates software to transform how distributed applications discover and consume data on-prem or on the cloud, and empowers end users to iterate faster and easier, creating new collaboration workflows among teams.

Want the presentations?