D2iQ, the enterprise Kubernetes provider for smart cloud-native applications, has released the latest version of its Kaptain AI/ML platform. According to the company, Kaptain AI/ML 2.1 is the first smart cloud-native platform to enable Nvidia GPU Container Catalog (NGC) containers to be launched directly from Kubeflow, empowering developers with pretrained, best-in-class GPU-optimized models for greater accuracy in production.

Kaptain AI/ML 2.1 features the first-ever seamless integration of Kubeflow and MLflow, giving users metadata tracking and visualization that enables improved performance models and the tracking of experiments directly from their notebooks. The integration means data scientists no longer have to choose between Kubefllow and MLflow.

The latest version also integrates seamlessly with DKP 2.3, enabling enterprises to standardize their infrastructure, running ML pipelines and other workloads on a single enterprise-ready platform.

The company added that Kaptain AI/ML is the only Kubeflow platform to eliminate all critical Common Vulnerabilities and Exposures in all components, highlighting security as a priority. In addition, Kaptain AI/ML 2.1 includes stronger identity provider integration. When combined with the military-grade security features in DKP, Kaptain AI/ML 2.2 provides an exceptionally secure AI/ML pipeline.

The new release also enables users to run Kaptain AI/ML workloads on Amazon Elastic Kubernetes Service (EKS) and Microsoft Azure Kubernetes Service (AKS), extending deployment options while further simplifying artificial intelligence (AI) and machine learning (ML) operations. Kaptain 2.1 continues strong support for air-gapped environments.

Additional key features of Kaptain AI/ML 2.1 include simplification of the user interface and management tasks. Customer-led enhancements, including Kaptain AI/ML’s simplified user interface, accelerate time-to-value and operational success for AL/ML workloads.

Kaptain AI/ML 2.1 is now generally available.