Kubeflow has seen wide interest from across industries as a technology to automate data science workflows, from data extraction to monitoring models in production. However, Kubeflow is composed of 30+ microservices that need to be deployed, configured, maintained, integrated and independently upgraded whenever there is a new release of one of the components.
In this webinar, we walk you through a few of the challenges of operating Kubeflow and how the latest technologies of software operations and lifecycle management can help, namely the most recent model-driven operators.