Snowflake has introduced a new way to handle dbt projects with Cortex Code. The tool automates model generation, testing and SQL optimization directly on the Snowflake platform. Engineers can now run these tasks from a single prompt or via the command line without switching between environments.
Cortex Code integrates with dbt by generating models based on existing schemas and business rules. It writes the necessary SQL files and configures tests automatically. The system also checks for performance bottlenecks in the generated queries and suggests optimizations. Users report faster iteration times when building data pipelines.
The automation covers the full dbt workflow. It creates staging models, transforms data into business-level aggregates and handles documentation generation. Tests are attached to each model to catch schema drift or data quality issues early. The CLI interface allows developers to push changes directly to Snowflake’s warehouse without manual uploads.
Early adopters include data teams at financial services and healthcare companies. One engineering lead said the tool cut their model deployment time from days to hours. Snowflake positions Cortex Code as a bridge between raw data and analytics-ready tables, aiming to reduce the gap between SQL developers and data consumers.
The feature is now available in public preview for Snowflake customers on the Enterprise edition and higher. Documentation and sample projects are accessible through the Snowflake UI.
Source: snowflake.com