Semantic layer

A shared, governed definition of a company's business metrics and entities (revenue, customer, churn) that people, BI tools and AI can all query consistently.

Definition

A semantic layer sits between raw data and the tools that use it. It defines what 'active customer' or 'net revenue' means once, in code, with the joins and filters behind it. Every report, dashboard and AI assistant then asks the semantic layer instead of writing its own query.

For AI, the semantic layer matters because language models are good at turning questions into queries but have no way of knowing your company's definitions. Pointing them at a governed semantic layer, or a knowledge graph of business entities, reduces wrong answers and keeps AI numbers consistent with finance's numbers.

Sources

  1. dbt Labs documentation, the dbt Semantic Layer