Arrikto Enterprise Kubeflow

Portable and Scalable ML Environment

One consistent Kubeflow environment from desktop to cloud.

Build ML models quickly on your laptop, GCP, or AWS with MiniKF, an all-in-one, single-node Kubeflow distribution containing Minikube, Kubeflow, Kale, and Rok. Deploy the same experience into production with Arrikto’s multi-node Enterprise Kubeflow machine learning operations offering. Move ML workflows seamlessly across with Rok Registry.

Portable and Scalable ML Environment

One consistent Kubeflow environment from desktop to cloud.

Build ML models quickly on your laptop, GCP, or AWS with MiniKF, an all-in-one, single-node Kubeflow distribution containing Minikube, Kubeflow, Kale, and Rok. Deploy the same experience into production with Arrikto’s multi-node Enterprise Kubeflow machine learning operations offering. Move ML workflows seamlessly across with Rok Registry.

Automated Machine Learning Workflow

Enable data scientists to easily create production-ready Kubeflow pipelines for MLOps.

Generate Kubeflow pipelines from ML code in any notebook with Kale. Start by tagging cells in Jupyter Notebooks to define pipeline steps, hyperparameter tuning, GPU usage, and metrics tracking. At the click of a button, create pipeline components and KFP DSL, resolve dependencies, inject data objects into each step, and deploy the data science pipeline. Or use the Kale SDK with your preferred IDE.

Reproducible ML Pipelines

Roll back instantly to any ML pipeline step for easy debugging and collaboration

Automatically snapshot pipeline code and data for every step with Rok, Arrikto’s advanced data management platform. Roll back to any machine learning pipeline step at it’s exact execution state for easy debugging. Collaborate with other data scientists through a GitOps-style publish/subscribe versioning workflow.

Reproducible ML Pipelines

Roll back instantly to any ML pipeline step for easy debugging and collaboration

Automatically snapshot pipeline code and data for every step with Rok, Arrikto’s advanced data management platform. Roll back to any machine learning pipeline step at it’s exact execution state for easy debugging.  Collaborate with other data scientists through a GitOps-style publish/subscribe versioning workflow.

Kubeflow Security and Access

Isolate users and their data with RBAC and fine-grain authorization controls.

Manage teams and user access via GitLab or any ID provider via Istio/OIDC.  Isolate user ML data access within their own namespace while enabling notebook and pipeline collaboration in shared namespaces.