How many data leaders have a semantic layer today?

44% of data and analytics leaders have already implemented a semantic layer, and another 48% plan to by 2027 — meaning roughly 92% of the market is moving on what Gartner calls a non-negotiable foundation.

How does Unwind Data help with the semantic layer?

Unwind Data designs and implements semantic layers across dbt Semantic Layer, Cube, LookML, Omni, Atscale, and custom builds — from semantic readiness assessments to ongoing managed services.

Gartner Data & Analytics Summit · London · 11–13 May 2026

Gartner's 2026 verdict on AI: the semantic layer is the non-negotiable foundation.

Q: What did Gartner say about semantics at the 2026 Data & Analytics Summit?

Distinguished VP Analyst Rita Sallam argued that agentic AI outcomes depend on semantic representations of data — the business meaning AI systems need to operate accurately at scale. Without it, AI agents hallucinate, introduce bias, and produce unreliable results.

Q: What is Gartner's 2027 prediction for semantics in AI?

Gartner predicts that by 2027, organizations that prioritize semantics in AI-ready data will increase agentic AI accuracy by up to 80% and reduce costs by up to 60%.

Across three days of keynotes in London, one signal came through louder than any other: only 1 in 5 AI investments show measurable ROI today, and the gap isn't the models — it's the semantic layer the models actually run on.

Book a semantic readiness call What is a semantic layer?

1 in 5

AI investments today show measurable ROI

Gartner, May 2026

+80%

Agentic AI accuracy gain by 2027 with semantics

Gartner forecast

−60%

Cost reduction by 2027 with semantics

Gartner forecast

92%

Of D&A leaders have or plan a semantic layer by 2027

44% today + 48% planned

The takeaways

What came out of the London summit

Across the May 2026 Gartner Data & Analytics Summit in London, analysts spent three days mapping the gap between AI ambition and AI ROI. The keynote that reframed the conversation came from Distinguished VP Analyst Rita Sallam: agentic AI outcomes depend on context — semantic representations of data, the business meaning AI systems need to operate accurately at scale.

Without it, agents hallucinate. They introduce bias. They produce unreliable results faster than humans can catch them.

The headline forecast: by 2027, organizations that prioritize semantics in AI-ready data will increase agentic AI accuracy by up to 80% and reduce costs by up to 60%.

Gartner's advice to D&A leaders, repeated across multiple sessions: establish a context layer as a core component of your data infrastructure. Schema-based models alone no longer suffice — they lack business meaning. Treat the layer like cybersecurity: strategic risk, executive accountability, board-level conversation.

Read the full Gartner announcement

Why this matters for you

The thing we've been building is now the thing Gartner calls non-negotiable.

Unwind Data has been implementing semantic layers since before the term started showing up in board decks. The Gartner announcement is validation — and a deadline. The companies winning with AI in 2027 aren't the ones buying better models. They're the ones funding the layer the models actually run on.

Semantic models AI can ground against

We define your metrics, dimensions, and business rules in one governed place — the layer that turns vague LLM prompts into accurate, auditable queries.

Tool-agnostic implementation

dbt Semantic Layer, Cube, LookML, Omni, Atscale, or a custom build. We pick the right fit for your stack, not the one we're paid to sell.

Governance the board can sign off on

Certified definitions, lineage, and access controls — so semantics becomes the cost-control and trust strategy Gartner is now calling non-negotiable.

From assessment to managed service

Start with a semantic readiness assessment, ship an MVP layer, then keep it healthy as your business and your AI stack evolve.

The Founder's take

What it looks like from inside the work.

Wesley Nitikromo, founder of Unwind Data, broke down the Gartner announcement on LinkedIn the day after the keynote. Rita Sallam — the Gartner analyst who delivered it — reposted it.

See the full post on LinkedIn

Wesley Nitikromo

Founder, Unwind Data · Reposted by Rita Sallam

Gartner put a number on it yesterday. 1 in 5 AI investments show measurable ROI.

“Semantics is the brain of AI.”

The high-performing AI organizations already show the pattern. Roughly 60% of their AI spend goes to foundations. Data architecture. Governance. Semantic context. The layer that makes the AI useful, not the AI itself.

44% of data and analytics leaders have already implemented a semantic layer. Another 48% plan to by 2027. Three-quarters of the market is now catching up on what Gartner is calling a “non-negotiable foundation.”

That moves semantics from a technical decision to a budget decision. From the data team's problem to the CFO's problem.

The companies winning with AI in 2027 are not the ones buying better models. They are the ones funding the layer the models actually run on.

What to do this quarter

Three honest next steps

Audit your AI spend

If less than 60% of it is going to data foundations, governance, and semantic context, you're budgeting against the high-performer pattern Gartner just published.

Pick a semantic layer

dbt Semantic Layer, Cube, Omni, Atscale, LookML — they all work. The wrong move is waiting for a perfect choice while your AI agents keep hallucinating.

Treat it as board-level

Sallam's framing: treat the context layer the way you treat cybersecurity. Strategic risk. Executive accountability. Not a backlog ticket.

Build the layer your AI actually runs on.

A 30-minute call. We map your current data foundations against the Gartner pattern and tell you, honestly, where the gap is.

Book a free call Explore our services