Skip to main content
1 min read ·

Semantic Layer

A semantic layer is an abstraction between raw data sources and consumers that defines business metrics, hierarchies, and relationships once — so every consumer (dashboards, AI agents, spreadsheets) reads the same governed logic regardless of access method.

Why This Matters

Without a semantic layer, the same metric gets calculated differently depending on who queries it and through which tool. Finance reports one revenue number. Marketing reports another. An AI copilot reports a third. Each reads from raw tables and applies its own interpretation. Humans reconciled this through institutional knowledge. AI cannot.

A governed semantic layer eliminates this at the infrastructure level — “revenue” means one thing, “gross margin ” applies one cost allocation, hierarchies roll up one way. Gartner now classifies semantic layers as “critical infrastructure.” The Open Semantic Interchange (OSI) specification was finalised in January 2026.

Where This Fits

Sits within Data Governance & AI Readiness as the mechanism that makes a dual-consumer data model operationally possible. Without it, governance is policy. With it, governance is infrastructure.

Let's go!

Expand your knowledge with our resources

Explore our comprehensive library of articles, guides, and tutorials to deepen your understanding of key concepts and stay up-to-date with the latest developments.

Book a free consultation