Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
This article describes Databricks Designated Services. Designated Services are features in Azure Databricks that process customer content according to geographies to manage data residency. It covers which features are Designated Services and how those features process data in various geographic locations.
Important
If a Designated Service is available in your workspace's Geo, then your content is processed in that Geo. Some Designated Services are not natively available in workspace Geos, and require cross-Geo processing be enabled on the workspace to use them. If you enable cross-Geo processing on your workspace, the Designated Service can process your content in other Geos. Importantly, your data is not stored outside of your workspace's Geo regardless of this setting.
Designated Services include:
- Databricks Foundation Model APIs
- Foundation Model Fine-tuning
- Agent Bricks
- Databricks AI-powered features
To learn more about Azure data residency, see Data residency in Azure.
EU Data boundary
Designated Services for workspaces within the EU Data boundary rely on EU data centers for processing.
Pay-per-token Foundation Model APIs availability
Foundation Model APIs allow you to access and query a curated set of state-of-the-art open models. Pay-per-token Foundation Model APIs are recommended for getting started with foundation models.
The following table shows the geographic areas that support pay-per-token workloads and where those curated models operate. Where these models operate is based on GPU availability, therefore some geographic areas might require that cross geography processing be enabled on workspaces. Those areas are listed in the table and where those workloads are routed when cross geography processing is enabled.
Note
These mappings might change for workspaces that have cross geography processing enabled.
See the models and regions available for pay-per-token workloads.
| Geographic area where your workspace is located | Azure Geo where Foundation Model APIs is hosted |
|---|---|
| United States | United States |
| Asia | United States |
| EU Data Boundary | EU Data Boundary |
| UAE North | EU Data Boundary |
| United Kingdom | EU Data Boundary |
| All other Geos | United States |
Provisioned throughput Foundation Model APIs availability
Foundation Model APIs allow you to access and query a curated set of state-of-the-art open models. Provisioned throughput Foundation Model APIs are for production foundation model workloads that require performance guarantees.
Where these models operate is based on GPU availability, therefore some geographic areas might require that cross geography processing be enabled on workspaces. Those areas are listed in the table and where those workloads are routed when cross geography processing is enabled.
Note
These mappings might change for workspaces that have cross geography processing enabled.
See the regions available for provisioned throughput workloads.
| Geographic area where your workspace is located | Geographic area where Foundation Model APIs is hosted |
|---|---|
| United States | United States |
| Asia Pacific | Australia |
| Australia | Australia |
| Brazil | United States |
| Canada | United States |
| EU Data Boundary | EU Data Boundary |
| India | Australia |
| Japan | Australia |
| Korea | Australia |
| UAE North | EU Data Boundary |
| United Kingdom | EU Data Boundary |
| All other Geos | Not available |
Foundation Model Fine-tuning
With Foundation Model Fine-tuning, you can use your own data to customize a foundation model to optimize its performance for your specific application.
The following table shows the geographic areas that support Foundation Model Fine-tuning and where those models operate.
| Geographic area where your workspace is located | Geographic area where Foundation Model Fine-tuning is hosted |
|---|---|
| United States | United States |
| All other Geos | Not available |
Agent Bricks
Agent Bricks provides a simple, no-code approach to build and optimize domain-specific, high-quality AI agent systems for common AI use cases.
The following table shows the geographic areas that support Agent Bricks.
| Geographic area where your workspace is located | Geographic area where Agent Bricks is hosted |
|---|---|
| United States | United States |
| All other Geos | Not available |
AI-based autocomplete availability
AI-based autocomplete provides inline code suggestions as you type in Databricks notebooks, the SQL editor, and the file editor.
The following table shows the geographic areas that support AI-based autocomplete and where those models operate.
| Geographic area where your workspace is located | Geographic area where AI-based autocomplete is processed |
|---|---|
| United States | United States |
| EU Data Boundary | EU Data Boundary |
| All other Geos | Routed to United States geography |
Databricks Assistant
Databricks Assistant is an AI-based companion pair programmer and support agent that makes you more efficient when creating notebooks, queries, and files.
| Geographic area where your workspace is located | Geographic area where Assistant is processed |
|---|---|
| United States | United States |
| EU Data Boundary | EU Data Boundary |
| Australia | Australia |
| India | India |
| All other Geos | Routed to United States geography |
AI-generated comments in Unity Catalog tables
You can use Catalog Explorer to add AI-generated comments for any table or table column managed by Unity Catalog. AI-generated comments provide a quick way to help users discover data managed by Unity Catalog.
| Geographic area where your workspace is located | Geographic area where AI-generated comments are processed |
|---|---|
| United States | United States |
| EU Data Boundary | EU Data Boundary |
| All other Geos | Routed to United States geography |
Mosaic AI Agent Evaluation availability
Agent Evaluation helps developers evaluate the quality, cost, and latency of agentic AI applications, including RAG applications and chains.
The following table shows the geographic areas that support Mosaic AI Agent Evaluation and where those models operate.
| Geographic area where your workspace is located | Geographic area where Mosaic AI Agent Evaluation is processed |
|---|---|
| United States | United States |
| EU Data Boundary | EU Data Boundary |
| All other Geos | Routed to United States geography |
AI/BI dashboards
AI/BI dashboards feature AI-assisted authoring, an enhanced visualization library, and a streamlined configuration experience so that you can quickly transform data into sharable insights. Dashboards use AI services to create visualizations from natural-language prompts and support companion Genie spaces.
| Geographic area where your workspace is located | Geographic area where Assistant is processed |
|---|---|
| United States | United States |
| EU Data Boundary | EU Data Boundary |
| Australia | Australia |
| India | India |
| All other Geos | Routed to United States geography |
AI/BI Genie
AI/BI Genie provides a way for business teams to interact with their data using natural language. It uses generative AI tailored to your organization's terminology and data, with the ability to monitor and refine its performance through user feedback.
| Geographic area where your workspace is located | Geographic area where Genie is processed |
|---|---|
| United States | United States |
| EU Data Boundary | EU Data Boundary |
| Australia | Australia |
| India | India |
| All other Geos | Routed to United States geography |