Microsoft and Snowflake: Simplified interoperability with no data movement

Data today lives everywhere—across apps, services, and clouds. Every department has its own analytics stack, AI tools, and preferences, and what used to be a manageable data landscape is now a distributed web of systems. But now, in the era of AI, bringing this data together has never been more important as we build agentic systems that need access to data across the organization. True interoperability—where platforms connect seamlessly, and data doesn’t have to move—is quickly becoming the key to unlocking value at scale.
That’s why Microsoft and Snowflake have been working side by side to make open, cross-platform integration effortless. Over the past 18 months, our collaboration has focused on one shared goal: helping customers connect Snowflake and Microsoft OneLake to access, analyze, and share data without duplication or complexity.
Built on open standards like Apache Iceberg and Parquet, this collaboration lets organizations use a single copy of data across both platforms and choose the right tool for every task. The result is a more flexible, efficient, and unified data experience—no matter where your data originates.
To learn more about how this interoperability works, check out our recent Microsoft and Snowflake: Delivering on the promise of openness and interoperability blog post.
Microsoft Ignite: Announcing enhanced interoperability between Microsoft and Snowflake
We’re excited to share new advancements that make the Microsoft–Snowflake integration even easier to use and more powerful.
We’ve added new, intuitive user interface (UI) experiences in both platforms to simplify setup and use. OneLake is adding a Snowflake-branded item in preview, allowing users to seamlessly access all Snowflake data within Microsoft Fabric without requiring further configuration. This means you can use any Fabric workload—analytics, AI, or visualization—directly on Snowflake data, without extra configuration.
Snowflake is also introducing new UI capabilities designed to let OneLake serve as the native storage location for your Snowflake data. This means all of your data can reside in OneLake, while taking advantage of Snowflake’s powerful engines.
Take a look at this new UI in action below and get started today.
How does this add to Microsoft’s existing interoperability?
We’ve already been able to deliver bidirectional data sharing between Snowflake and OneLake, for seamless interoperability between our platforms without data duplication. Customers can already write Snowflake tables directly to OneLake, access Apache Iceberg tables using OneLake shortcuts, and read OneLake tables from Snowflake—all without duplication or complex setup.
What we’ve already delivered:
- General Availability
- Automatic translation of Iceberg metadata to Delta Lake metadata for use with all Microsoft Fabric engines.
- Shortcut Snowflake Iceberg data (in Azure, Amazon S3, or GCS) directly into OneLake.
- Preview
- Native storage of Snowflake Iceberg data in OneLake.
- Automatic conversion of Fabric data into Iceberg format for seamless use in Snowflake.
- New OneLake table APIs that work with Snowflake’s catalog-linked database feature.
And with the new UI now rolling out, we are making the existing interoperability easier to implement for your teams.
Looking ahead to unified, cross-platform data access and management
Looking ahead to 2026, our goal is to make all these capabilities generally available, so that even your most mission-critical workloads can take advantage of unified, cross-platform data access and management.
But beyond our existing interoperability, we are committed to continue removing barriers between our platforms, so you have full optionality for your data projects.
Still have questions about the integration?
Watch the recent Ask me Anything: Fabric and Snowflake Interoperability webinar where experts from Microsoft OneLake and Snowflake answered top questions on how to most effectively use these platforms together.