How WHOOP bridges dbt models and Snowflake semantic views
WHOOP built an open-source bridge between dbt and Snowflake Semantic Views, and deployed 80 of them in four months.
Overview
WHOOP built snowflake-semantic-tools (SST), an open-source package letting teams define semantics in YAML alongside dbt models and deploy to Snowflake Semantic Views. Covers why semantic consistency matters for AI, how SST converts YAML into Semantic Views, and how a small team built this without dedicated engineering resources.
What you'll take away
- 01
Snowflake Semantic Tools, an open-source CLI, generates Snowflake Semantic Views directly from dbt so a metric gets defined once instead of drifting between dbt and the warehouse.
- 02
Four months in, WHOOP deployed 80+ Semantic Views in use by 50+ people across 5 teams, built on roughly half of one engineer's time across two quarters.
- 03
The team paired the CLI with six custom Cortex Code skills committed into the project, giving analysts a guided path to the semantic layer instead of learning its internals from scratch.
Session speakers
William Tsu
WhoopMadhura Pharande
WHOOPSophia Scaglioni Melegari
WHOOP