# Databricks

Databricks lakehouse. Dex profiles Unity Catalog tables and authors dbt models with DBU-aware cost guards.

## Why connect Databricks?

Connecting Databricks means your team stops exporting it by hand and starts working with it directly inside Dex.

### What you get

1. Query live data in Databricks without exporting it or standing up a new pipeline
2. Keep dashboards and dbt models in sync with Databricks instead of a stale nightly copy
3. Give teams governed, self-serve access to Databricks under the same RBAC and audit trail as everything else

## How it works

Explore, transform, maintain: the Dex loop on Databricks.

### Point Dex at Databricks

Install Dex in Claude Code by running the following commands:

- `/plugin marketplace add exmergo/exmergo-agent-plugins`
- `/plugin install dex@exmergo`
- `/dex:explore` against Databricks. Dex connects with your Databricks credentials directly, so they, and your raw data, never enter the model's context.

### Explore before touching anything

`/dex:explore` profiles the schema it finds in Databricks and proposes what to model before writing a single line of dbt.

### Author governed dbt models

`/dex:transform` writes and maintains the dbt models and tests that turn raw Databricks data into trusted, reusable tables in your project.

### Stay inside a cost guard

Dex enforces a cost guard on every query it runs against Databricks, so exploration never turns into a surprise bill.

### Catch drift automatically

`/dex:maintain` flags it when Databricks's schema changes and proposes the fix, instead of letting your dbt project silently drift out of sync.

## Other warehouses

Dex brings the same loop and the same guardrails to each of these:

- BigQuery
- Snowflake
- Redshift
- ClickHouse
- PostgreSQL
- Trino
- DuckDB

## Need a connector we don't support yet?

Tell us about your data source and we'll let you know as soon as the connector is available.
