Last week we announced that Domino is now fully Kubernetes native.
This is great news for data science teams and IT organizations building modern DS platforms, for two reasons:
- First, Domino now installs on any infrastructure — any cloud or on prem, even in your own k8s cluster — so you can future-proof your data science infrastructure and integrate seamlessly into your existing stack.
- Second, we aren't merely using k8s to install our platform services — we went deep: we rebuilt our whole compute grid to leverage k8s to distribute data science workloads, including model training jobs, interactive notebook sessions, hosting apps and deploying production models. With Domino, all your data science workloads get lift from k8s, automatically. This makes data scientists more efficient, saving them dev ops hassle. It also reduces IT support burden, and reduces infrastructure costs by efficiently distributing workloads.
After being the first platform to containerize data science workloads in 2014, I’m proud we're again leading the industry in bringing the power of modern infrastructure to data science teams. K8s is an amazing foundation for the next wave of innovations coming soon.
Btw, if you like k8s and data science, join our rapidly growing engineering team!
Editorial note: Nick's post originally appeared on LinkedIn. For more details about how Kubernetes is an orchestration framework that can bridge the gaps that have existed between data scientists and IT, see his article "Kubernetes: The key ingredient IT needs to accelerate today’s data science" in Venturebeat.