Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More
Microsoft is pushing the bar on simplifying data efforts for enterprises. Today, at its ongoing Build 2024 conference, the company announced that it is expanding its partnership with Snowflake to enable bi-directional data access for joint customers.
The move comes with support for Apache Iceberg and will power interoperability between Microsoft Fabric – the company’s end-to-end data analytics platform – and Snowflake’s data cloud, saving customers from the hassle of keeping multiple copies of data at different locations. It will also drive major cost savings in the long run, the companies said.
“We are enabling this experience by accelerating our adoption of open standards, collaborating with industry-leading partners, and making it easier for customers to rapidly benefit from the innovations we are bringing to the market,” Arun Ulagaratchagan, corporate vice president of Azure Data at Microsoft, said in a statement.
Understanding Microsoft and Snowflake’s interoperability
Last year at Build, Microsoft unveiled Fabric, a unified platform with the ability to handle an organization’s data and analytics workloads end-to-end, right from data integration and engineering to visualization. The platform drew widespread attention in the industry with its use of OneLake, a data lake that used open formats – Apache Parquet and Delta Lake – to bring together the data from Microsoft’s ecosystem and external sources, including third-party applications.
“The reason we made the move to support open formats, rather than proprietary formats, is because customers value the idea that their data is completely open all the time. And because it’s Apache Parquet and Delta Lake, hundreds of tools that work with the formats natively work with our customer’s data,” Ulagaratchagan told VentureBeat.
Now, as the next step in this work, Microsoft is expanding its engagement with Snowflake and adding support for another open format: Apache Iceberg. This way, Snowflake will support OneLake as a native data store, allowing joint customers to store and access their data as a single copy in the Iceberg format on Fabric. Meanwhile, the Iceberg data in OneLake will become accessible via Snowflake.
Before this, the two platforms did not interoperate and users had to build pipelines and maintain copies of data on each platform, which took a lot of time and financial resources.
With bi-directional storage and access, Microsoft says users will be able to integrate their Snowflake data with apps in its ecosystem via Fabric. This could be anything from Teams and Excel to Power BI and Azure AI Studio. Similarly, data stored in Fabric’s OneLake can be extended to Snowflake to power cross-cloud AI, applications via Cortex, sharing and collaboration, and advanced analytics among many other workloads.
“What customers want is to be able to say I have my data. I control my data and then I can mix and match technologies and engines to the specific use case and that is where we’re coming together to announce,” Christian Kleinerman,” Christian Kleinerman, EVP of Product at Snowflake, told VentureBeat.
Roll out later this year
While the news is out, the interoperability experience between Snowflake and Microsoft Fabric is yet to go live. The companies plan to work on this integration over the coming months and launch it in preview later this year.
“We’re seeing massive interest in this from customers we’ve spoken to. It’s under active development. We don’t have anyone testing it yet, but there’s a huge line of enterprises that are just super excited to take advantage of it,” Arun added.
Microsoft Build runs from May 21 to May 23, 2024.