Arrikto announced today the immediate availability of its enterprise-grade distribution of Kubeflow on Microsoft Azure, completing its multi-cloud distribution on the top three public clouds. The company also announced the immediate, free availability of MiniKF for Kubeflow 1.3, dramatically simplifying the workflow for data scientists and MLOps teams to build, test, and deploy models faster than any MLOps platform on the market.
A new Kubeflow community survey shows Kubeflow’s popularity in production environments soaring 300% over the past year. To support these deployments at scale, Arrikto announced today that it has released Enterprise Kubeflow for Microsoft Azure to complete its support for all three major cloud providers – Google Cloud Platform, Amazon AWS, and now Microsoft Azure. Organizations deploying Enterprise Kubeflow on Microsoft Azure can accelerate their models into production 30-times faster than traditional approaches to MLOps. Organizations deploying on Arrikto Enterprise Kubeflow on Azure can:
- Manage cluster lifecycles using GitOps
- Replicate ML code & data snapshots across Kubeflow clusters
- Test the same ML code across training and production environments
- Share notebooks and datasets across clusters and between users
- Scale their ML environments to multi-node
- Provide enterprise-grade security, including IAM integration, shared namespaces, credential management, repo pub/sub, and integration with external services
- Create an AutoML workflow with the click of a button. Start from a dataset, define a task, then discover, train, and optimize a model from inside your notebook automagically.
The company also announced that it has updated its popular MiniKF – the easiest and fastest way to get started with Kubeflow. Now available with Kubeflow 1.3, MiniKF includes key improvements that simplify model development through deployment. including:
- Expanded monitoring of KFserving components, including predictor, transformer, and explainer
- Streamlined Katib hyperparameter tuning with trial performance visualization
- Simplified TensorBoard configuration and accessibility to metrics to improve model accuracy and performance
- Easily create and delete volumes in your namespace with the new Volume Manager
- Quickly setup tuning with the new Kale hyperparameter wizard UI
- Added Kubeflow notebook support for TF 2.0, PyTorch, VS Code, and RStudio (in addition to TensorBoard)
- Extended ML framework support for TorchServe predict and PyTorch Captum, PMMLServer, and LightGBM
MiniKF comes bundled with Kubeflow 1.3, Kale for automating MLOps workflow, and Rok for ML data management. MiniKF is available to run on a desktop via Vagrant or in a single-node distribution on Google Cloud Platform and AWS marketplace. A new Codelab tutorial covering model building through serving is also available to get started quickly. To download MiniKF and get started with the Codelab, visit arrikto.com/get-started.
Arrikto enables MLOps teams to accelerate machine learning models to market 30-times faster than traditional ML platforms. As a leading Kubeflow contributor, Arrikto provides automated workflows, reproducible pipelines, consistent deployment from desktop to cloud, and secure access to data. Arrikto’s Enterprise Kubeflow is available as a multi-node distribution on AWS, GCP, and Azure, and is the preferred MLOps platform used in production today by many Fortune 500 enterprises. The company has over 300 customers across 17 countries and is venture backed by Unusual Ventures and Odyssey VP.