How WHOOP bridges dbt models and Snowflake semantic views

Analytics teams struggle with business logic scattered across tools. Meanwhile, AI accelerates analysis but can amplify inconsistencies without a governed semantic layer. Defining semantics in YAML enables metrics and relationships to live in version control. 

At WHOOP, we wanted semantics defined in YAML alongside our dbt models and deployed directly as Snowflake Semantic Views. No tools existed for this workflow, so we built one using dbt and AI code assistants, without dedicated software engineering resources. The result is snowflake-semantic-tools (SST), an open-source Python package that converts version-controlled YAML into Snowflake Semantic Views. 

This session covers why semantic consistency matters, how SST works, and how a small analytics team built a production-grade solution that keeps metric definitions aligned across every tool that touches the data.