canonical: https://jentic.com/apis/azure.com/azure-databricks-management

# Microsoft Azure DatabricksClient

Jentic publishes the only available OpenAPI specification for DatabricksClient, keeping it validated and agent-ready. The Azure Resource Manager API for Databricks exposes 7 operations across 4 paths to provision, list, update, and delete Azure Databricks workspaces. A workspace is the top-level Databricks tenant on Azure that holds clusters, notebooks, jobs, and data; this management API handles its lifecycle including SKU and managed resource group configuration. Use it to stand up new Databricks environments for analytics teams, list existing workspaces by subscription or resource group, and decommission them when projects end.

## For AI agents

Provision and manage Azure Databricks workspaces through 7 ARM operations on the Microsoft.Databricks resource provider, covering create, list, update, and delete with SKU and managed resource group configuration.

## Scope

Does not run notebooks, manage clusters, jobs, or secrets inside the workspace, or query data - use for Azure Resource Manager workspace lifecycle management only.

## Capabilities

- Create new Databricks workspaces with a chosen SKU (standard, premium, trial) in any supported region
- List Databricks workspaces in a subscription or scoped to a resource group
- Retrieve a workspace's full configuration including managed resource group ID and provisioning state
- Update tags and SKU on an existing Databricks workspace
- Delete a Databricks workspace and trigger cleanup of its managed resource group
- Enumerate Microsoft.Databricks operations available to the caller for permission checks

## Use cases

### Provision Analytics Workspaces on Demand

Create Databricks workspaces from CI or platform automation to give new data teams an isolated analytics environment. The Workspaces_CreateOrUpdate endpoint accepts the SKU, location, and managed resource group name; Azure provisions a managed resource group containing the underlying VM scale set, network, and storage. Provisioning typically completes in 10-15 minutes and the workspace URL is returned in the response.

Example prompt: Create a Databricks workspace named 'analytics-prod' in resource group 'rg-analytics' with premium SKU in westeurope and tag costCenter=data-platform

### Workspace Inventory and Cost Reporting

List every Databricks workspace in a subscription to feed a cost dashboard. The Workspaces_ListBySubscription and Workspaces_ListByResourceGroup endpoints return SKU, location, tags, and managed resource group, which combine with the Cost Management API for spend per workspace. This is the standard pattern for a centralised data platform team running dozens of workspaces.

Example prompt: List all Databricks workspaces in subscription /subscriptions/{id}, sort by SKU, and report each workspace's location and tags

### Workspace Decommissioning

Delete Databricks workspaces when projects close or teams disband. Workspaces_Delete removes the workspace and triggers cleanup of its managed resource group, releasing all underlying compute and storage. The operation is asynchronous; poll the returned status URL for completion. Pair with Cost Management to confirm spend stops shortly after.

Example prompt: Delete the Databricks workspace 'analytics-old' in resource group 'rg-analytics' and poll until the operation completes

### AI Agent Data Platform Operations

An AI agent supporting a data platform team can manage Databricks workspaces through Jentic. The agent searches by intent ('create azure databricks workspace'), Jentic returns the Workspaces_CreateOrUpdate operation with its schema, and the agent executes the call with managed Azure credentials. The same flow handles updates, listing, and deletion across the platform's full footprint of workspaces.

Example prompt: Search Jentic for 'create an Azure Databricks workspace', load the Workspaces_CreateOrUpdate schema, and execute it with the user-supplied workspace name, SKU, and location

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| GET | /subscriptions/{subscriptionId}/providers/Microsoft.Databricks/workspaces | List Databricks workspaces in a subscription |
| GET | /subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/Microsoft.Databricks/workspaces | List Databricks workspaces in a resource group |
| DELETE | /subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/Microsoft.Databricks/workspaces/{workspaceName} | Delete a Databricks workspace |
| GET | /providers/Microsoft.Databricks/operations | List Microsoft.Databricks operations |

## Key resources

- **Workspaces** — Create, list, update, and delete Azure Databricks workspaces
- **Operations** — List Microsoft.Databricks resource provider operations available to the caller

## Why Jentic

- **Setup:** Wiring the Azure Databricks management API by hand means registering an Azure AD app, handling MSAL bearer tokens against management.azure.com with the user_impersonation scope, and polling the Azure-AsyncOperation header for workspace create and delete yourself. Through Jentic you install once, import Databricks Management from the API Directory, store the Azure service principal credential once, and your agent calls it.
- **Permission scoping:** Databricks puts the workspace name in the URL path (/subscriptions/{id}/resourceGroups/{rg}/providers/Microsoft.Databricks/workspaces/{workspaceName}), so a rule can pin your agent to one workspace or resource group. You choose the operations it may call, so a destructive one like Workspaces_Delete is not included unless you add it alongside read operations such as Workspaces_Get or Workspaces_ListByResourceGroup.
- **Credential handling:** Your Azure service principal credential is stored once, encrypted, by your own Jentic One instance and injected as a scoped ARM bearer token at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'create an Azure Databricks workspace' or 'list Databricks workspaces', and Jentic returns the matching ARM operation with its input schema so the agent supplies the SKU and location without reading the reference docs.

## Related APIs

- **Data Factory Management** — Orchestrate Databricks notebooks as part of a wider ETL pipeline.
- **HDInsight Management** — Open-source Hadoop, Spark, and Kafka clusters on Azure instead of managed Databricks.
- **Storage Management** — Provision Azure Storage and ADLS Gen2 accounts that Databricks reads and writes.

## FAQ

### Why is there no official OpenAPI spec for DatabricksClient?

Microsoft Azure does not publish an OpenAPI specification. Jentic generates and maintains this spec so that AI agents and developers can call DatabricksClient via structured tooling. It is validated against the live API and kept up to date. Get started with Jentic One, the self-hosted execution layer.

### What authentication does the Azure Databricks management API use?

It uses Azure Active Directory OAuth 2.0 against https://management.azure.com/ with the user_impersonation scope. The caller needs Contributor or a custom role with Microsoft.Databricks/workspaces/* permissions on the resource group. Jentic stores service principal credentials encrypted and issues short-lived bearer tokens at execution time.

### Can I create premium-SKU workspaces with this API?

Yes. The Workspaces_CreateOrUpdate body accepts a sku.name field with values 'standard', 'premium', or 'trial'. Premium enables features like role-based access control, secret management, and JDBC/ODBC endpoint access controls.

### What are the rate limits for the Databricks management API?

Azure Resource Manager applies subscription-level read and write throttling (around 1,200 requests per hour). Long-running operations such as workspace creation and deletion return 202 Accepted with an Azure-AsyncOperation header you poll for completion.

### How do I create a workspace through Jentic?

Run pip install jentic and search for 'create azure databricks workspace'. Jentic returns the Workspaces_CreateOrUpdate operation; load its schema and execute it with the subscriptionId, resourceGroupName, workspaceName, and a body specifying location, sku, and managedResourceGroupId.

### Does this API run notebooks or manage clusters inside the workspace?

No. This is the ARM management plane only - it provisions the workspace itself. Cluster, job, and notebook operations use the Databricks REST API at https://{region}.azuredatabricks.net/ with personal access tokens or Azure AD tokens for the workspace.

### Can I limit what my agent is allowed to do with the Azure Databricks Management API?

Yes. Jentic One runs self-hosted, so your own rules decide which operations and credentials the agent may use. Because the workspace name sits in the URL path, you can pin the agent to a single workspace or resource group, and you pick the exact operations it may call. You can grant read-only access such as Workspaces_Get and Workspaces_ListByResourceGroup while leaving out a destructive operation like Workspaces_Delete unless you deliberately add it.