Domino Data Lab, provider of the Enterprise MLOps platform, is making upgrades to its model monitoring capabilities, allowing companies to place greater trust in the models they deploy. This and other enhancements—including Domino Model Monitor (DMM) support for AWS, GCP, and Azure— are part of Domino 4.6.
New in this release, DMM can automatically compute data drift and model quality across billions of daily predictions, more than 100 times the scale of previous versions, according to the vendor.
This is made possible by the new Domino Elastic Monitoring Engine, a technology advancement that scales model monitoring capacity infinitely to support the most demanding monitoring requirements.
Customers gain broad visibility and control across a limitless number of models with hundreds or even thousands of features to help them avoid unwanted outcomes across their business, according to the vendor.
“Successful model-driven businesses have thousands of models driving critical business processes,” said Nick Elprin, co-founder and CEO of Domino Data Lab. “By helping our customers detect drift faster and across more of their models, Domino is reducing risk while freeing up data scientists’ time for new research.”
DMM now supports a broad range of deployment options including major cloud platforms (AWS, GCP and Azure), as well as on-premises deployments.
DMM can also now connect directly to Amazon S3 and Hadoop Compatible File Systems (HCFS), which include Azure Blob Store, Azure Data Lake (Gen1 and Gen2), Google Cloud Storage, and HDFS.
This powerful new capability allows production data to be analyzed directly from its native storage location, eliminating the need to copy data from place to place with the potential to introduce version control, security, and other issues.
Domino 4.6 also introduces easily deployable access to Ray.io and Dask distributed compute frameworks, which along with existing Spark support, builds on Domino’s vision for allowing data scientists to work the way they want with the tools and infrastructure they need. Distributed compute allows data scientists to process large volumes of data for machine learning and other complex mathematical calculations such as deep learning.
Additionally, DevOps Teams can select the best framework for the job at hand—without being limited to the single choice provided by other platforms —and then dynamically provision and orchestrate workloads directly on the infrastructure backing a Domino instance without the need for DevOps skills or IT intervention.
This added support also makes it easy for data science teams to leverage pre-packaged machine learning libraries and scale the data handling capacity of their applications with little or no changes to existing models.
Other features in Domino 4.6 include:
- Domino certification of Amazon EKS. Amazon’s managed Kubernetes offering is now part of the “Certified” family of highly secure, comprehensively tested, and performance-optimized Domino offerings to streamline installation and upgrades.
- Git Repository Creation & Browsing with CodeSync. Users can now create Git repositories, then browse and edit their files easily—all without leaving the Domino environment.
- Read/Write Domino Datasets. Significant improvements in the user experience of Domino Datasets allow them to read, write, and share large amounts of data seamlessly across the entire Domino platform.
- Single Sign-On. Additionally, customers can leverage Single Sign-On (SSO) across the entire Domino Enterprise MLOps Platform for a seamless experience, from model development to deployment to monitoring.
For more information about this release, visit www.dominodatalab.com.