A SQL-first data transformation framework that brings software engineering practices (version control, testing, documentation, modular models) to analytics workflows — the 'T' in ELT pipelines.
dbt (data build tool) lets analytics engineers write SQL transformations as version-controlled models, run them as a DAG, test them (schema and data tests), and document them automatically. Models reference other models with Jinja templating, dbt resolves dependencies, and the warehouse executes the SQL. Adopted across nearly every modern data stack on Snowflake, BigQuery, Redshift, and Databricks. The de facto standard for the transformation layer between raw data lake/warehouse and BI tools.
Replacing 2,000 lines of unversioned, untested SQL in scheduled scripts with 60 dbt models in Git — full lineage, automated tests, generated docs.
dbt is what made 'analytics engineering' a real discipline — applying software practices to SQL pipelines dramatically improved data reliability and team scalability.
Need help implementing this in your business?
Get Started