Glossary
What is a semantic layer?
Last updated
Definition
A semantic layer is a shared set of definitions for business terms and metrics, such as revenue or active customer, kept in one place so every dashboard, report and AI tool calculates them the same way.
Why teams build one
Without a shared definition, finance and sales report different revenue. A semantic layer fixes that by defining the metric once.
Where it lives
- In a BI tool, such as LookML in Looker or a Power BI semantic model.
- In the warehouse, as views, which every tool can read.
- In a separate metrics service that sits between the warehouse and tools.
How exelens uses the idea
exelens does not ask you to build a new model. It reads the warehouse views you import, so the view is the definition. If you already run Looker or Power BI, see exelens vs Looker and exelens vs Power BI. Book a demo.
Frequently asked questions
- Do I need a semantic layer to use AI analytics?
- You need clear definitions. A semantic layer is one way to hold them. Curated warehouse views are another, and often the simplest.
- Is a view a semantic layer?
- A view can carry a metric definition, which does part of the job. A full semantic layer adds relationships, reusable metrics and access rules across many tools.
See Semantic layer in practice
A short walkthrough with the founder: connect read-only, ask a real question, pin the answer.