This week’s “From BI to AI” update is a little different from the usual. Snowflake Summit occurred June 8-10, bringing a slew of announcements related to Snowflake’s new features, and many Snowflake partners timed their own related announcements in sync with the Summit.
Snowpark and Java UDFs
On Tuesday, June 8, Snowflake launched Snowpark, their “developer experience.” Data scientists, data engineers, and developers can build in Java or Scala within Snowpark, and then execute their workloads directly within Snowflake.
Also on the coding side, Snowflake announced support for Java UDFs (user-defined functions) within Snowflake, allowing customers to import their custom code and business logic to Snowflake. Both Snowpark and Java UDFs within Snowflake are currently in private preview, with public preview coming soon.
Snowflake also announced the Snowpark Accelerated Program, where partner vendors can access Snowflake technical experts and be provided with additional exposure to existing Snowflake customers.
Snowflake Partner Announcements
Numerous Snowflake partner vendors followed up with their own announcements on Wednesday, June 9.
Snowflake can now be used as a data source within Amazon SageMaker Data Wrangler. This integration allows data prep for machine learning in SageMaker to occur in Snowflake.
Alteryx announced a deeper integration of Alteryx with Snowflake. Alteryx Designer is now directly available on Snowflake; data prep, data blending, and automated analytics processing are pushed down into Snowflake for better performance and scalability. Joint Alteryx-Snowflake customers can also augment existing data sources with those available on the Snowflake Data Marketplace. Current Snowflake customers have access to a free trial of Alteryx within their Snowflake account.
Dataiku debuted their Snowflake integration with Snowpark and Java UDFs. Dataiku-Snowflake users will be able to push computation down to Snowflake, so that data prep and scoring can happen within Snowflake.
DataRobot’s new Snowflake integration also joins DataRobot with Snowpark, growing the existing DataRobot-Snowflake partnership. Data prep tasks from Zepl (a recent DataRobot acquisition) can be pushed into Snowflake for feature engineering, while providing a preconfigured environment for model development within Snowpark. DataRobot’s Java Scoring Code also pairs with Snowflake Java UDFs to enable DataRobot models to do scoring within Snowflake.
Domino Data Lab inaugurated its Snowflake partnership this week with Snowpark integration as well. Joint Domino-Snowflake customers will be able to build data pipelines within Snowpark, and execute MLOps workflows from Domino within Snowflake.
Building on its 2020 Partner of the Year status, Informatica announced tighter integrations between its Intelligent Data Management Cloud and Snowflake, offering support for Java UDFs for joint Informatica-Snowflake customers and advancing its mass-ingestion ELT capabilities. Users will be able to transform, cleanse, and govern data from a wide variety of enterprise applications automatically, en masse, on its way to ingestion in Snowflake.
Finally, Talend revealed Talend Trust Score for Snowflake. This new capability will allow joint Talend-Snowflake users to verify data quality within Snowflake, using Snowpark and Java UDFs.
(Sidenote: the acquisition of Talend by Thoma Bravo is proceeding apace; Thoma Bravo has begun the tender offer to acquire all outstanding ordinary shares and American Depository Shares of Talend.)
If you would like your announcement to be included in Amalgam Insights’ weekly data and analytics roundups, please email email@example.com.