What is a semantic layer?
A semantic layer is a layer of business meaning placed over raw tables and columns, so that cryptic field names map to the concepts people actually use — like "revenue" or "active customer." iDBQuery's semantic layer lets it understand obscure schemas and answer questions in your real-world terms.
Raw databases are rarely self-explanatory. A column might be called `amt_net_fx` or a table `tbl_cust_03`, and the relationships between them live only in someone's head. A semantic layer is the translation between that raw structure and the language people use to ask questions.
iDBQuery builds and applies a semantic layer so it knows that a particular column is the figure you mean by "net revenue," that two tables join on a specific key, and which records count as an "active" customer.
A semantic layer delivers:
- Plain-language understanding — ask for "churned accounts last quarter" and it knows which fields and filters that implies.
- Consistency — the same business term resolves to the same definition every time, so numbers agree across answers.
- Cryptic-schema rescue — even badly named or undocumented databases become answerable.
iDBQuery's semantic layer is provenance-aware: definitions can come from your own annotations, a business glossary, data dictionaries, or be inferred automatically from the schema and sample data, with your explicit definitions winning. This is what lets a non-technical user get a correct answer from a database they've never seen, and what keeps the meaning of a metric stable across an entire team.
Updated 2026-06-22