The challenge of translating raw enterprise data into actionable insights has long plagued businesses, a problem that often overshadows the capabilities of sophisticated data warehouses and visualization tools. This fundamental disconnect is precisely what Omni, a company founded four years ago, aims to resolve, and investors are now committing significant capital to its approach. The company recently finalized a $120 million Series C funding round, spearheaded by Iconiq, pushing Omni’s valuation to an impressive $1.51 billion.
At its core, Omni develops what is known as a semantic layer—a governed intermediary positioned between a company’s raw data and the various entities, human or artificial, querying it. This layer functions as a dynamic rulebook, meticulously defining key metrics like revenue, establishing access controls for sensitive figures, and standardizing calculation methodologies. This foundational technology is already in use by a diverse client roster, including BambooHR, Guitar Center, Checkr, Mercury, Pendo, and Heidi AI. Notably, BambooHR alone leverages Omni to serve over 100,000 users within its operational framework.
The genesis of Omni traces back to a reunion of its three co-founders, Colin Zima, Jamie Davidson, and Chris Merrick, all graduates of Princeton University. Their paths converged again after Google’s 2020 acquisition of Looker, their previous employer, for $2.6 billion. Zima, who now serves as Omni’s CEO, previously held the roles of chief analytics officer and vice president of product at Looker, bringing considerable experience in data analytics and product development to his new venture.
The landscape for semantic layer solutions is not without competition. Established players and emerging innovators are all vying for market share. OpenAI, for instance, introduced Frontier in February, explicitly positioning it as “a semantic layer for the enterprise that all AI coworkers can reference.” Similarly, tech giants like Snowflake and Databricks have integrated their own semantic layer offerings directly into their existing enterprise-grade platforms. Omni’s strategic response to this bundling threat lies in its architectural design. The company contends that legacy providers would face substantial re-engineering efforts to replicate what Omni has meticulously built from the ground up. Matt Jacobson, a partner at Iconiq, draws a parallel between Omni’s current advantage and Snowflake’s early competitive edge over Amazon’s Redshift, suggesting a similar structural superiority.
Colin Zima characterizes the present technological moment as a genuine inflection point for the industry. He notes that the demand for such a solution has always been latent, but the necessary tools were simply not available until now. This sentiment appears to be validated by Omni’s recent performance metrics; the company’s Annual Recurring Revenue (ARR) has nearly quadrupled over the past year, and it achieved profitability for the first time just last month. Currently, Omni employs approximately 200 individuals across its offices in San Francisco, Dublin, and Sydney, reflecting its expanding global footprint.
Zima views the burgeoning AI wave not as a disruptor but as a significant accelerant for Omni’s growth. He explains that as more enterprises integrate AI agents to interact with their data, the imperative for a robust, governed semantic layer becomes increasingly critical. This perspective positions AI as a tailwind, rather than a headwind, for Omni’s business model. Iconiq’s Jacobson echoes this sentiment, emphasizing the sheer scale of the market opportunity. He suggests that this opportunity far surpasses the scope of traditional business intelligence as previously conceived, noting that the broader business intelligence software market is projected to reach approximately $47 billion by 2025, with the semantic layer sub-segment anticipated to grow at an annual rate of 30% through 2031, according to the Futurum Group. The pace of enterprise adoption has also accelerated dramatically, with implementation measured in days and weeks rather than months or years.
