Semantic layer
A semantic layer maps raw data fields to business terms like revenue or churn, so every tool and person queries the same definition.
What is a semantic layer?
A semantic layer sits between raw data tables and the tools that query them, mapping technical fields to business terms your team actually uses, like revenue, active customer, or churn. It defines a metric once, including how it's calculated and which tables it draws from, so a dashboard, a spreadsheet, and a data scientist's query all return the same number for the same term.
How does a semantic layer differ from a context layer?
A semantic layer standardizes structured metrics for analytics: it defines what a number means and how to compute it, mainly for BI tools and dashboards. A context layer serves a broader purpose: it's the infrastructure that gives AI agents access to a company's knowledge, structured facts, relationships, and documents alike, kept current and permissioned for retrieval at any time. A semantic layer can define what revenue means. A context layer is what lets an agent know which deals make up this quarter's revenue, who owns them, and whether that information is still current.
How does a semantic layer work?
A semantic layer is built by defining metrics and dimensions once, in code, on top of the underlying data tables, then exposing those definitions through a single interface that every connected tool queries against instead of writing its own SQL. When a metric's definition changes, it changes everywhere it's used, instead of requiring an update in every dashboard that referenced the old version. Qontext's context repository takes the same principle further for AI agents: instead of standardizing only numeric metrics, it keeps a company's structured knowledge, facts, relationships, and their sources, current and consistent across every agent that reads from it.