dbt Labs

dbt (data build tool) is the open-source standard for data transformation in SQL, combining SQL development with real-time validation and state intelligence to deploy data models quickly and reliably. The platform unifies raw data (extract via Fivetran, Airbyte) and transformations (SQL dbt) directly in the warehouse, eliminating external movements and tools. The Fusion engine (new) improves team productivity building data models locally or in cloud. dbt offers advanced capabilities like model dependency management (automatic DAG), quality tests (key existence, uniqueness, references), versioning/deployment via dbt Cloud, and recently the Semantic Layer to expose business metrics defined once to all BI tools (Power BI, Looker, Tableau). Plans range from free Developer (1 seat, 3k models/month, 14d pro trial) to Starter ($100/seat/month, 5 developers, 15k models) and Enterprise custom (unlimited seats/models, 30 projects, dbt Canvas and dbt Mesh).

dbt does not provide a French-language interface and remains dominated by English-speaking technical users. The open-source dbt-core remains free and self-hostable, ideal for teams with dedicated infrastructure; dbt Cloud (managed) adds orchestration, testing, and continuous deployment. The REST API and native integration with warehouses (Snowflake, BigQuery, Databricks) simplifies automation. dbt excels for data engineers comfortable in SQL and code management; the learning curve is low for SQL-proficient profiles but very steep for non-data-savvy users. The true cost depends on data engineer count: $100/seat is affordable for small teams but escalates quickly with more developers.