Databricks has announced a broader set of OpenSharing and Marketplace updates for the AI era, expanding sharing beyond traditional datasets and making it easier to collaborate across platforms, clouds, and partner applications.
Key highlights
- OpenSharing builds on Delta Sharing: It is the next evolution of Delta Sharing and adds open, vendor-neutral sharing for AI assets like agent skills, AI models, and unstructured data.
- Broader interoperability: OpenSharing expands the existing ecosystem with support for Iceberg-compatible access, helping providers reach more recipients across tools and platforms.
- Apps on Marketplace: Customers can now discover, install, and run partner-built apps directly inside their Databricks workspace, where apps run in a secure environment and inherit Unity Catalog governance.
- SecureConnect simplifies external sharing: SecureConnect introduces a Databricks-managed proxy for storage access, so providers can do a one-time setup instead of configuring network access for every recipient.
- Hybrid and on-prem data are part of the story: With OpenSharing and the storage ecosystem, organizations can connect on-premises and private-cloud data to Databricks without copying the underlying data first.
In the full post, you’ll see how Databricks is extending sharing from tables to a broader set of data and AI assets, while also making it easier to run partner apps and simplify secure collaboration across clouds and environments.
Read all the post here 👇:
