Semantic for dbts
The dbt semantic layer defines business metrics once, in YAML, using semantic models, entities, dimensions and measures through MetricFlow. Tools and AI assistants then query governed definitions like revenue instead of rewriting SQL, which keeps numbers consistent across dashboards, notebooks and LLM-based question answering.
Stage: Deploy and Serve · Part of the dbt roadmap · Last reviewed 19 November 2026
What to learn in Semantic
- Define a semantic model with entities, dimensions and measures
- Create a metric such as revenue or conversion rate in YAML
- Query a metric by dimension and time grain with MetricFlow
- Explain why central metric definitions prevent conflicting dashboard numbers
- Describe how an AI assistant can use governed metrics instead of raw SQL
Semantic interview questions
What problem does a semantic layer solve?
Different teams often compute the same metric with slightly different SQL, producing conflicting numbers. A semantic layer defines each metric once, and every tool queries that definition, so results match across dashboards, reports and applications.
What are entities, dimensions and measures in a semantic model?
Entities are keys that link semantic models, such as customer or order. Dimensions are attributes you slice by, like country or date. Measures are aggregations over columns, such as a sum of amount, that metrics build on.
Where Semantic fits in the dbt roadmap
Semantic belongs to the Deploy and Serve stage (advanced level). Production.