Arrikto raises 10 million to advance its ML platform

Arrikto’s technology makes deploying and managing Kubeflow simpler, allowing data scientists to manage it using the tools they are already familiar with. It also helps create a portable environment for data science to enable data sharing and data versioning across teams and clouds. You can quickly build models on your laptop, GCP, or AWS with MiniKF, an all-in-one, single-node Kubeflow distribution containing Minikube, Kale, and Rok.

With Arrikto, data scientists can easily generate production-ready pipelines for MLOps. It allows you to create pipeline components and KFP DSL, resolve dependencies, inject data objects into each step, and deploy the pipeline swiftly. You can also tag cells in Jupyter to define pipeline steps, hyperparameter tuning, GPU usage, and metrics tracking.

It is also possible to roll back to any pipeline step at its exact execution state to debug. You may also collaborate with other data scientists through a GitOps-style publish versioning workflow.

Arrikto also allows you to isolate ML data access within the user’s namespace while allowing notebook and pipeline collaboration in shared namespaces. Isolating users and their data with RBAC and fine-grain authorization access ensure secure access.