What is the semantic layer and how does it help?
The semantic layer is a meaning layer iDBQuery builds on top of your raw schema so it understands what cryptic table and column names actually represent. It records definitions, business terms and relationships — so when you ask about "revenue" or "active customers," iDBQuery maps those words to the right fields.
Raw database schemas are rarely self-explanatory. Columns get named `cust_typ` or `amt_net_2`, and the same concept lives under different names in different systems. The semantic layer is how iDBQuery bridges that gap between human language and your actual data.
It captures:
- Annotations on tables and columns — what each field really means, merged from multiple sources (your own notes, a data dictionary, documentation, learned patterns and the schema itself), with the most authoritative source winning.
- A business glossary so shared terms like "net revenue," "churn" or "active student" resolve to a consistent definition everywhere.
- Relationships between tables and across sources, so the same entity is recognised even when systems name it differently.
iDBQuery can bootstrap a lot of this automatically when you connect a source, and you can refine it. The payoff is twofold: the AI writes far more accurate queries because it knows what your words mean, and your whole team shares one agreed vocabulary. Without it, an AI is guessing at your schema; with it, iDBQuery is reasoning over data it genuinely understands.
Updated 2026-06-22