Google Cloud has announced Cloud AI Platform Pipelines to make the process of deploying robust repeatable machine learning pipelines, easy. The AI PlatformPipelines will also provide a way to monitor, audit, version track and deliver enterprise-ready, “easy to install, secure execution environment for your ML workflows.”
But what’s new?
The new AI Platform Pipelines will provide a user with a push-button installation via the Google Cloud Console.With Pipeline, users will have new enterprise features for running ML workloads, including pipeline versioning, automatic metadata tracking of artifacts and executions, Cloud Logging, visualization tools. It will also enable seamless integration with many Google Cloud managed services Dataflow, Serving, AI Platform Training, BigQuery and Cloud Functions.
Google on its blog post: “The beta launch of AI Platform Pipelines includes a number of new features, including support for template-based pipeline construction, versioning, and automatic artifact and lineage tracking.”
To start using AI PlatformPipelines, you will have to navigate to AI Platform> Pipelines, then click on NEW INSTANCE.From here, you can choose if you want to use an existing GKE cluster or create a new one. To create a new cluster, you will have to check a box to access to any Cloud Platform service from the pipelines.