putilov_denis - stock.adobe.com

DataRobot integrates AI modeling tools with Snowflake

The AI vendor and cloud data warehouse vendors' latest collaboration will enable joint customers to develop and deploy AI models with fewer production lags.

AI vendor DataRobot and cloud data warehouse vendor Snowflake revealed a new integration to ease developers' ability to build AI models in Snowflake's Snowpark platform.

DataRobot said at the Snowflake Summit 2021 virtual conference today that it will provide joint customers a preconfigured environment to ease the development, training and deployment of AI models in Snowpark, a developer experience environment.

The new integration comes in the wake of DataRobot's acquisition and integration on May 11 of Zepl's modeling tools into its AI platform. Zepl's notebook enables data scientists to more easily incorporate their code and apps into AI models. It also comes after DataRobot in February expanded its existing partnership with Snowflake by integrating the AI vendor's Feature Discovery tool with the Snowflake data warehouse.

Snowflake's Snowpark

Meanwhile, at this week's conference, Snowflake expanded Snowpark -- now in preview -- with added support for Java functionality and an enhanced SQL API for querying the cloud data warehouse.

With the additional level of integration between DataRobot and Snowpark, developers will be able to write AI models in multiple languages and execute various types of data workloads such as extract, load, transform, or ELT, optimally from within Snowflake's Data Cloud, according to DataRobot.

The tighter integration between DataRobot's Zepl and Snowflake could minimize the production gaps that hamper AI development. Currently, much AI production runs into bottlenecks due to disparate machine learning models and data, DataRobot said.

More integration between the two fast-growing vendors could benefit developers -- but there are already existing links that make that possible.

Existing Integration Points

"Enterprises that use both Snowflake and DataRobot will appreciate this partnership because the integration will make accessing data for AI easier," said Mike Gualtieri, an analyst at Forrester Research.

However, it is not particularly compelling for prospects because DataRobot can easily integrate with other data platforms and Snowflake can easily integrate with other AI platforms, Gualtieri added.

Also, Snowflake has been entering into numerous third-party partnerships in its effort to expand its cloud data warehouse technology into a wider cloud data platform. For example, on Tuesday analytics vendor ThoughtSpot rolled out its second big integration with Snowflake.

This may be strategic for DataRobot and Snowflake but I am not sure it matters much to enterprise buyers.
Mike GualtieriAnalyst, Forrester

DataRobot and Snowflake first entered into a partnership in 2018. In 2020, the vendors amped up the relationship when Snowflake Ventures invested in DataRobot as part of a $320 million Series F funding round in order to help accelerate product and technology integration between the two platforms.

As part of that investment, the companies also agreed to further accelerate product and go-to-market strategies. 

At the summit, the two vendors have scheduled a virtual hands-on lab to demonstrate how Snowflake's data cloud integrates with DataRobot's Automated Machine Learning platform.

"This may be strategic for DataRobot and Snowflake, but I am not sure it matters much to enterprise buyers," said Mike Gualtieri, an analyst at Forrester. "Both companies have been on a partnership tear lately."

DataRobot in 2018 introduced a specialized partner program. In June 2020, the AI vendor partnered with Boston Consulting Group and acquired some of its software assets.

Next Steps

DataRobot acquires Algorithmia to further MLOps goal

Snowflake targets generative AI with new capabilities

Dig Deeper on AI infrastructure

Business Analytics
Data Management