Why data scientists shouldn’t need to know Kubernetes

Recently, there have been many heated discussions on what the job of a data scientist should entail [1, 2, 3]. Many companies expect data scientists to be full-stack, which includes knowing lower-level infrastructure tools such as Kubernetes (K8s) and resource management. This post is to argue that while it’s good for data scientists to own the entire stack, they can do so without having to know K8s if they leverage a good infrastructure abstraction tool that allows them to focus on actual data ...

Submit