The next generation of enterprise applications will increasingly be AI/ML models applied to accelerate existing processes or solve new problems such as accelerating drug discovery and development in life sciences. Kubeflow is an open source reference architecture for AI/ML platform initiated by Google and contributed by several IT platform infrastructure leaders in the industry such as IBM, Redhat, Cisco, Dell, AWS for on-prem and hybrid deployment of AI/ML.
Often the data available for developing AI/ML models cannot be moved to the cloud and/or requires to be on a hybrid installation depending on the purpose. One Convergence as a member of the Kubeflow community has developed DKube, an MLops platform using Kubeflow ref architecture primarily for on-prem and hybrid use. DKube is a supported distribution of Kubeflow, which requires a Kubernetes platform underneath allowing the applications to be cloud native while being on-prem. Just like Kubeflow, having a supported Kubernetes distribution facilitates the enterprise deployment of Kubeflow.
MLOps is more than just having a platform for model training and deployment. It is a practice to have a workflow- from rapid prototyping with data and then training a model, optimizing a model through hyper parameters and further experimentation, then publishing models that fit best, and then deploying from the published catalog. Later a deployed model may have to be retired or replaced with a new better model based on the feedback of its quality of prediction. That whole loop is MLOps. Without such discipline AI models never make it production or cannot do the job in the field.
With Nutanix Karbon customers get that supported Kubernetes. Along with Nutanix Objects and Files as the foundation underneath Dkube brings AI/ML model development and deployment wherever Nutanix customers are, wherever their data is - onprem or in the cloud. Between the support infrastructure of Nutanix and One Convergence customers will get a fully supported AI/ML engine. Just like Nutanix's core platform it runs on Nutanix's own appliances as well as third party servers available from Dell, HPE, Cisco, Supermicro, Lenovo and others.
“Next generation workloads such as MLOps demand a performant, scalable and resilient infrastructure that can deliver business outcomes rapidly. Nutanix cloud-native stack including Karbon and Objects enables our partner OneCovergence to deliver a complete solution with its DKube MLOps platform to our joint customers with on-prem and in-cloud deployment flexibility.” - Priyadarshi Prasad, VP Products, Nutanix
“We are delighted to be partnering with Nutanix to bring MLOps workloads on top of Nutanix stack. So much data is available on Nutanix stack and is ripe for applying AI/ML models to it. And with Nutanix now on AWS and Microsoft Azure gives customers a true hybrid platform to run AI/ML models wherever it makes most sense, wherever the data moves” - Prasad Vellanki, CEO, One Convergnece.
DKube is now a Nutanix Ready AI/ML product with installation and operation within the same day on top of Nutanix infrastructure. With DKube customers will get the most cost effective, most innovative AI/ML platform in the industry.
For more information about DKube please visit https://www.dkube.io or send a note to email@example.com.
DKube is owned by One Convergence Inc., based in San Jose, CA.
Your data and LLM models reside in private cloud but GPU resources are spread across clouds.
Protecting intellectual property in your corporate documents and LLM models, both inside and outside your organization.