dbt is a tool for transforming raw data in a data warehouse into trusted data products. You write business logic as SQL select statements, called models, and dbt turns them into tables and views, manages execution order, and runs tests and documentation alongside the code. dbt Core v2.0, the open-source release of the new engine written in Rust, is still in alpha and is released under the Apache License 2.0. You run it yourself from the command line; no Python runtime is needed.
dbt is developed by dbt Labs, which merged with Fivetran in June 2026 and operates together with it as a company; both are US companies. dbt Core itself does not store data: it runs on your own machine or CI environment and sends SQL to the warehouse you choose, so where data is stored is your decision, including in the EU. The dbt distribution also contains components under a separate dbt licence, and an optional hosted dbt platform service is available.