Skip to main content

Databricks - Private Integration

J
Written by Johan Åkerman
Updated this week

Note:

This integration is in beta and only available for selected enterprise customers. Please reach out to your engagement manager for early access.

Introduction

The integration allows read operations to Databricks and allows users to query and visualize data from Databricks inside of Sana Agents.

Operation type

Supported

Read

Write

Common use-cases:

Analyze data inside of Databricks with natural language, see example result below for the query “give me progression in temperature for Havana and plot it” for a user who connected Sana to their database with weather data.

Type of integration

Private: Set up per user and mirroring the underlying individual access rights.

Availability

Integration type

Free tier

Team tier

Enterprise tier

Private

Shared

Centralized

Scope and permissions

Scope

Purpose

openid

Enables OpenID Connect; required to receive an ID token during authentication.

email

Grants access to the user’s email address (included in the ID token or user info).

profile

Grants access to basic user profile info (name, username, picture, etc.).

offline_access

Enables refresh tokens, allowing long-lived access without re-authentication.

sql

Databricks-specific scope that permits access to the SQL API (e.g., SQL warehouses).

all-apis

(Optional) Grants access to all Databricks APIs where authorized.

Integration set-up

  1. Select Databricks in the integrations tab

2. Click the “Connect (OAuth2)” button

3. Fill in workspace connection details

4. Click “Connect just for me”

5. Sign in with your Databricks account and accept scopes

6. Add a view by either selecting and importing an existing view in the sidebar or define your own view

6.1.1 Define your own view by filling in the title, description and the SQL definition before clicking the “Fetch and describe columns” button, see example below. Add

6.1.2 Add custom description for the columns if wanted, then click “Continue”.

6.2.1 Select a view in your available views and click “import Selected Views” at the bottom

6.2.2 Review the important view, add column descriptions if wanted and click “Continue”.

7. You’re now all set!

Known limitations

  • Databricks Genie is not supported.

FAQ

Q: What is a view?

A: A Databricks view is a virtual table defined by a saved SQL query. Sana can only see and query the views you explicitly share with it.

Q: Why should I add column descriptions if they are not required?

A: Adding brief descriptions gives Sana context about each column, so it can pick the right fields and return more accurate results.

Data handling & privacy

Sana AI is fully committed to data security and privacy. All data accessed by Sana AI is encrypted both in transit and at rest. Sana does not train any underlying language models on your data, ensuring the privacy of your information. Sana AI is ISO 27001 certified; and SOC 2 and GDPR compliant, and adheres to the highest standards of data security.

Did this answer your question?