Glossary
Context layer
Definition
A context layer is the business knowledge that AI systems draw on to interpret data correctly: definitions, rules, relationships and terminology, kept in one place so every AI tool uses the same understanding of the business.
What it means in practice
Large language models know a great deal about the world but nothing specific about your company. A context layer fills that gap. It tells an AI assistant what "active customer" means for you, which revenue figure finance reports, and how one department's terms map to another's.
The term overlaps heavily with "semantic layer". A traditional semantic layer focuses on translating business terms into SQL; a context layer emphasises the wider knowledge AI needs to reason, including knowledge that lives in documents and people's heads. A governed, certified context layer is often called a contextual semantic layer.
Example
Asked "How many active customers did we have last quarter?", an AI assistant with a context layer applies the company's definition (three or more orders in 90 days, excluding test accounts) and its fiscal calendar, instead of guessing.