- Business intelligence company Omni raised $120 million in new funding, per an exclusive Fortune report published April 24, 2026.
- Co-founder Colin Zima previously co-founded Looker, the analytics platform Google acquired for $2.6 billion in 2020.
- Omni’s semantic layer technology gives AI agents consistent, business-contextualized definitions of enterprise metrics—preventing the inconsistent query results that occur when models access raw database schemas directly.
- The raise accelerates Omni’s positioning as data infrastructure for AI-native enterprise deployments as organizations move from pilots to production.
What Happened
Omni, a business intelligence startup that builds semantic layers over enterprise data warehouses, raised $120 million in new funding, according to an exclusive published by Fortune on April 24, 2026. The company was co-founded by Colin Zima, who previously co-founded Looker before its acquisition by Google for $2.6 billion in 2020, alongside fellow Looker alumni. Per Fortune, Zima described the problem Omni is built to solve as one where enterprise AI systems querying business data without a governed semantic layer produce inconsistent answers—because the same metric can be computed differently depending on who wrote the query.
Why It Matters
The enterprise AI bottleneck has increasingly shifted from model capability to data infrastructure. When large language models are used as data-querying agents inside organizations, the absence of consistent metric definitions means a question like “what was last quarter’s churn rate” can return different numbers depending on which table, filter, or join the model constructs. The problem echoes the pre-Looker era of BI, when spreadsheet proliferation led to conflicting figures across departments—a dynamic Omni’s founders experienced firsthand. The semantic layer segment has grown considerably since 2023, with companies including dbt Labs, Cube, and AtScale competing in overlapping areas of the market.
Technical Details
Omni’s platform sits between a company’s cloud data warehouse—including Snowflake, BigQuery, and Databricks—and downstream consumers, whether human analysts or AI agents. The system encodes business logic—table joins, aggregation rules, filters, and metric definitions—into a versioned, governed catalog that any authorized query consumer must pass through. This layer eliminates what practitioners call “metric drift”: the divergence that occurs when different teams or tools independently construct queries for the same business concept. Omni exposes the semantic layer via a query API compatible with LLM-based analytics tools and agent orchestration frameworks, allowing AI systems to request resolved metric results rather than raw SQL generation against undocumented schemas.
Who’s Affected
Data engineering and analytics teams at mid-market and enterprise companies are the primary stakeholders. Organizations that have deployed AI assistants or autonomous agents for internal analytics—generating reports, answering business questions, or running data workflows—face the most direct pressure to standardize metric definitions before those deployments can scale reliably. Competing platforms including Looker (Google Cloud), Tableau (Salesforce), and open-source tooling in the dbt ecosystem face the same structural demand as enterprise buyers seek unified data semantics that serve both human and AI consumers from a single governed source.
What’s Next
Per the Fortune report, Omni intends to direct the $120 million toward engineering expansion and deeper integrations with AI agent frameworks. Zima has consistently positioned Omni as foundational data infrastructure for enterprise AI—a framing that puts the company in competition with both legacy BI vendors and newer entrants in the data catalog, observability, and metrics-layer categories as production AI deployments continue to surface data reliability as a primary constraint.
