The founders who built Looker into a $2.6 billion company just raised $120 million to solve one of the most stubborn problems in enterprise AI: getting agents to work with real business data without making things up.

Omni Analytics announced its Series C on April 23, 2026, led by ICONIQ with participation from GV, Theory Ventures, First Round Capital, and Redpoint Ventures. The round values the company at $1.5 billion (more than double its March 2025 valuation), and comes as revenue has already tripled year-to-date in 2026.

The founding team of Colin Zima (CEO), Jamie Davidson, and Chris Merrick all worked at Looker before Google acquired it. Zima served as Looker’s chief analytics officer and VP of product. After the acquisition, the three reconnected with a specific conviction: the AI era would need a governed translation layer between raw enterprise data and the agents and tools querying it.

That layer is what Omni builds.

What Omni Actually Does

When a business deploys an AI agent to answer questions about revenue, churn, or pipeline coverage, the agent has to query data somewhere. If it hits raw database tables directly, it faces a minefield of ambiguity. What counts as a “closed deal”: signed contract, payment received, or something else? Which regions are included in EMEA? How is “active user” defined this quarter versus last?

Without clear answers baked into the data layer, agents guess. And they often guess wrong.

Omni’s semantic layer acts as what the company calls a “governed context graph,” a living rulebook that defines business metrics, enforces permissions, and standardises calculations across every tool that touches the data. Instead of each AI agent or BI tool interpreting the raw data independently, every query runs through a single, agreed-upon definition layer.

Practically speaking, this means AI agents using Claude, ChatGPT, Cursor, or VS Code can query enterprise data through Omni and get answers that match what a human analyst would produce. The governance and logic sit in one place, not replicated (and inevitably corrupted) across dozens of downstream tools.

The Numbers Behind the Round

Omni’s growth tells a clear story about enterprise demand:

• 4x year-over-year revenue growth as companies consolidate legacy BI tools and accelerate AI adoption
• Revenue already tripled year-to-date in 2026
• Reached profitability for the first time in the month before the round closed
• Integrates directly with Snowflake, Databricks, and Google BigQuery, the main data warehouse platforms running enterprise analytics today

The $120M round includes a $30M employee tender offer, letting early team members and employees realise some value before any liquidity event.

ICONIQ led the round. GV (Google Ventures) also participated, which is notable given Google owns Looker, the very business these founders built. That’s either an endorsement of Omni’s approach or a hedge against it, possibly both.

Why the Semantic Layer Matters More Now Than Ever

Two years ago, semantic layers were primarily a BI problem. Data teams built them to keep finance and sales from arguing over whose revenue number was right.

Today, with AI agents querying business data in production environments, the stakes are higher. A wrong definition doesn’t just produce a bad dashboard. It produces a wrong decision made by an autonomous system, possibly without a human reviewing it.

This is why venture capital is flooding into companies that sit between enterprise data and enterprise AI. The problem is not compute or model capability. The problem is trust. Business leaders will not deploy autonomous agents on their data if those agents can produce credible-sounding but factually wrong answers about the business.

Omni’s pitch is that a governed semantic layer is the missing trust layer for enterprise AI. It is not a middleware curiosity. It is the infrastructure that makes agentic analytics safe to deploy at scale.

What This Means for Business

If you are running data analytics in your organisation and thinking about where AI agents fit in, this round is worth understanding. Here is what it signals:

The “connect AI to your data” problem is not solved yet. Most businesses that have tried plugging an AI assistant into their data warehouse have run into inconsistent answers. The semantic layer approach (enforced definitions, permissions, and metric governance) is emerging as the practical fix.

Your data quality work is not wasted. It is the foundation. Companies that have already invested in clean, well-defined metrics and governance frameworks are best positioned to benefit from AI analytics. The businesses that skipped that work will pay for it now.

BI consolidation is accelerating. Omni’s growth partly comes from companies stripping out legacy BI tools and replacing them with a unified platform that serves both human analysts and AI agents from the same governed layer. If you have Tableau, Looker, Power BI, and three other BI tools running in parallel, you are paying for the problem Omni is solving.

The race to own enterprise data context is on. Snowflake, Databricks, Salesforce, and now well-funded startups like Omni are all competing to be the authoritative definition layer for AI queries. Which platform your organisation anchors to for data governance will matter a lot over the next two years.

For data teams specifically, the semantic layer is becoming the most strategic asset you can build. It is not just about making dashboards consistent. It is about making your entire AI agent stack trustworthy.

---

Enterprise DNA helps organisations build the data skills and infrastructure that make AI actually work in production. Whether you are upskilling your team on data foundations or looking at AI-native business services, the starting point is always the same: clean data, clear definitions, and people who understand both.

If you’re deciding where to start with agents, start here. The free Working With Claude field guide walks through the ecosystem, Claude Code, and a real rollout plan. Get your copy.