Glossary
Contextual semantic layer
Definition
A contextual semantic layer is a semantic layer that also captures business context, such as why metrics exist, which rules and exceptions apply, and how terms differ across teams, so both people and AI interpret data the same way.
What it means in practice
A traditional semantic layer maps business terms to tables and generates SQL. A contextual semantic layer adds the knowledge around those terms: governed metric definitions, data models and relationships, business rules, and knowledge from documents, BI reports and subject-matter experts.
This extra context is what lets AI assistants and agents answer business questions reliably, explain how they got an answer, and stay consistent across tools. Codd AI builds contextual semantic layers with AI and keeps people in the loop to certify them.
Example
Beyond defining "net revenue", a contextual semantic layer records that finance excludes intercompany transactions, that the definition changed in Q4, and that sales uses "bookings" for a related but different number.