KazantsevAlexander - Fotolia
DataRobot expanded its partnership with Snowflake by integrating its Feature Discovery tool with the Snowflake cloud data warehouse.
Feature Discovery, which has been part of the DataRobot enterprise AI platform since 2019, enables users to automate the time-consuming process of feature engineering. With a visual relationships editor, users can set relationships between chosen data sets and carefully pick which data to include when calculating new features.
Using those defined relationships, Feature Discovery can automatically discover, test and create new features while providing a full lineage of how and where the features were generated.
The tool's new integration with Snowflake will enable users to perform data preparation operations in Snowflake, enabling them to access more data from Snowflake while cutting down on data movement and speeding up performance.
"This integration is valuable, as it empowers businesses to simplify the creation and implementation of features for machine learning and accelerate adoption and scalability," said Ritu Jyoti, program vice president for AI research at IDC. "Overall, it helps in unlocking data and feature silos with powerful predictions and actionable intelligence."
The integration comes after DataRobot revealed in December that Snowflake Ventures contributed to DataRobot's $320 million Series F funding round. The funding boosted DataRobot's valuation to more than $2.8 billion.
Alongside the investment, DataRobot entered into a partnership with Snowflake to deepen user experience with enhanced product integration and a go-to-market strategy, Jyoti noted.
Ritu JyotiProgram vice president, AI research, IDC
Responding to questions from SearchEnterpriseAI, a DataRobot spokesperson noted that DataRobot has a number of joint customers with Snowflake, although the spokesperson declined to say how many.
Currently, while DataRobot works with other cloud storage providers, it does not have a similar Feature Discovery integration with other cloud data storage vendors.
DataRobot said that, pending customer demand, it would consider adding support for this optimization to other data sources but doesn't yet have a timeline for that.