Write once, analyze anywhere: Omni + the dbt Semantic Layer

If you've ever spent an afternoon debugging a dashboard only to realize the "Active Users" definition in your BI tool doesn't match the one in your data warehouse, you know the pain of semantic drift. And now, as more teams rely on AI-powered tools to explore and summarize data, that inconsistency only compounds—AI can only be as trustworthy as the metrics it’s built on. It’s the exact problem the dbt Semantic Layer was built to solve: define your metrics centrally, in code, and query them consistently across your entire stack.

Today, we're thrilled to celebrate a massive milestone in that mission: building on top of its existing dbt integration, Omni has officially launched its first-class integration with the dbt Semantic Layer.

This is a big deal for analytics engineering teams because it allows you to reuse the logic you already have - wherever it is, and saves you a lot of repetitive work and debugging. Omni has always shared our philosophy of bringing software engineering best practices to BI, and this native integration takes that alignment to the next level.

For example, our team might define a core metric like ARR once in dbt, then explore it instantly in Omni, whether that’s in a dashboard, an ad hoc query, or an AI-driven workflow that summarizes performance trends for leadership.

We use Omni internally at dbt Labs to explore our own business metrics, including AI-powered analysis, so we’ve seen firsthand how valuable it is when governed definitions are available everywhere teams ask questions.

Here is a look at how it works, why this integration will save you time and headaches for all your traditional workflows, and how it will help your team as it accelerates AI adoption.

How it works: Your dbt metrics, native to Omni

On top of Omni's existing dbt integration, this new release deeply understands your semantic definitions. By pointing Omni at your dbt semantic layer project, your central definitions are automatically mapped directly into the Omni data model.

Here’s what that looks like in practice:

• Metrics: Your Simple, Ratio, and Derived metrics in dbt automatically map to corresponding views and measures in Omni. No recreating the math.
• Dimensions and Entities: dbt dimensions map directly to Omni dimensions, while your dbt entities map to Omni relationships, automatically building the correct join logic behind the scenes.
• Zero context switching: The descriptions, labels, and metadata you painstakingly curated in your dbt .yml files appear right in the Omni UI. When business users explore data, they have full context without ever leaving their workflow.

To get started, you just need to enable the dbt semantic layer integration in your Omni connection.

A massive step for open semantics (and the OSI spec)

At Coalesce 2025, we talked a lot about the future of interoperable semantics. Recently, we published the first version of the Open Semantic Interchange (OSI) specification alongside partners like Snowflake, Databricks, and Salesforce.

Our shared goal with OSI is to create a vendor-neutral, extensible model for representing metrics, dimensions, and relationships so they can be interpreted consistently across any tool or AI application. No more vendor lock-in; no more recreating business logic in five different platforms.

Omni's native integration is exactly the kind of workflow the OSI vision is built to enable. It proves that when you give teams more flexibility, they win. You author your business logic once in dbt, and your downstream tools—like Omni—simply know how to read it, safely and consistently. It’s a beautifully DRY approach to analytics.

Try it out

We’re excited to see the Omni team bring this integration to life. We’ve already experienced the benefits of Omni internally, and we know analytics engineers are going to love the seamless experience of governing metrics in dbt and exploring them in a best-in-class BI platform.

Ready to stop rewriting your metric definitions? Check out the Omni documentation to get started with the dbt Semantic Layer integration today.