Peter Dalbhanjan

This session takes you through the journey of enterprise AWS customer iRobot in adopting Kubeflow to run its machine learning (ML) workflows. It includes a discussion of evaluation criteria for Kubeflow and Kubeflow pipelines and how these tools help continuously improve the iRobot Genius AI feature.

In this hands-on workshop, learn how to build and run machine learning models using open-source software. We cover creating a Kubernetes cluster using Amazon Elastic Kubernetes Service (Amazon EKS) and accelerated EC2 instances; training, optimizing, and deploying an ML model to your cluster; and building a pipeline to automate the entire process.

Apache Spark is a popular execution framework for running machine-learning workloads. Since the beta release of spark-operator in early 2019, more users are determined to move their big data analytics workloads to Kubernetes.