SQL Developers Automate DBT without Code!

Karrots offers a command line interface environment to develop DBT. It’s exactly what SQL developers do today on their desktop, except that it’s entirely in the cloud. Our DBT access to the data warehouse is simple, secure: there are no long-term passwords or access tokens. But often SQL developers are not programmers and struggle to automate their DBT projects once tested and completed.

In Karrots, once a SQL developer commits their DBT project changes to git, the backend detects those changes and automatically processes them into an Airflow DAG automation job. This automation creates DAG nodes that correspond to the project’s models and tests. The SQL developer simply adds a few structured comments to the project that indicate the Airflow run schedule. When you look at the Airflow DAG for the DBT project, you see the DBT steps as graph nodes. When the DAG runs, the graph nodes run like any other DAG: in specified order with retry and reporting logic.