Follow Datanami:
April 20, 2021

Dataiku Gets Closer to Snowflake

Dataiku today revealed that Snowflake has invested an undisclosed amount in the data science platform provider. This comes several months after the two companies rolled out a joint offering that allows Dataiku customers to perform machine learning tasks in the Snowflake environment.

Dataiku is one of a number of startups seeking to bring AI and machine learning into the enterprise computing world. The New York City company develops software that manages and automates various aspects of the data science process, from data prep and visualization to feature selection and ML model tracking.

Snowflake, of course, has become one of the most popular cloud data warehouses in recent years. Like other cloud data warehouses, Snowflake’s takes a lot of the pain out of configuring and running a scale-out analytical database that allows customers to focus on their SQL queries instead of infrastructure.

With exabytes of data flowing into Snowflake, it’s no wonder that customers want to run data science, ML and AI workloads there, too. However, while Snowflake offers some of the data prep and engineering capabilities that data science teams will need, it really caters more to the data analysts than the data scientists. This is one of the reasons why Snowflake is partnering with the leaders in the data science arena.

Last year, Snowflake rolled out a new data science partner program that featured closer integration with the enterprise AI environment of Dataiku, in addition to those from DataRobot, H2O.ai, and Amazon SageMaker. Snowflake has also taken an equity stake in DataRobot, along with ThoughtSpot.

As part of that program, Dataiku and Snowflake collaborated to debuted Dataiku Online with Snowflake, which is a pre-integrated SaaS offering that lets Snowflake customers instantly bring customer data into the Dataiku platform for machine learning and data analysis.

While Snowflake ostensibly was designed to run SQL data analytics at-scale against big data sets, it’s becoming more of a general data platform, and that means machine learning workloads can run there too.

“It comes down to efficiency, security, and governance–having your data all in one place makes it far easier to get from data to insights without wasting hours manually prepping and securing data sets across different environments,” a Dataiku spokesperson says. “Many of our customers are large global enterprises, so the virtually unlimited scale of Snowflake’s databases is very attractive when using Dataiku’s machine learning suite.”

Last fall, Snowflake announced its Snowpark DataFrame API can not only run using SQL, but also Java, Python, and Scala. The ability to run Java functions created the opportunity for Dataiku to run machine workloads in Snowflake, the spokesperson says.

Dataiku declined to name the amount that Snowflake Ventures invested. The company announced a $100 million Series D round of funding last August, bringing the total funding to $246.8 million, per Crunchbase. The comanies currently have more than 80 joint customers, a 50% increase over the course of several months, the company says.

Related Items:

Dataiku’s Collaborative AI Platform Attracts New Investors

Snowflake: Not What You May Think It Is

Snowflake Extends Its Data Warehouse with Pipelines, Services

 

Datanami