Denodo joins open data initiative to tackle AI interoperability

Data management company Denodo has joined the Open Semantic Interchange (OSI), an open-source initiative aimed at solving one of enterprise AI’s most persistent headaches: fragmented and inconsistent data definitions across tools and platforms.

According to the company's press release, the initiative is led by Snowflake and brings together partners across business intelligence, data governance, AI, financial services, and manufacturing. Its goal is a common, vendor-agnostic specification that standardises semantic metadata — the business meaning attached to data — so that metrics and definitions remain consistent across dashboards, notebooks, and machine learning models.

“A universal, vendor-neutral semantic standard is critical for enterprises to accelerate innovation and unlock greater business value from AI and analytics,” said Kijoon Lee, Vice President of Strategic Alliances at Denodo. Lee added that OSI would enable interoperability between Denodo’s semantic layer and other technologies, “helping customers drive measurable business outcomes.”

Snowflake framed the effort as foundational infrastructure for the next wave of AI development. “Unlocking the full potential of data and AI requires a common foundation, and the Open Semantic Interchange is the critical step in building that bedrock,” said Josh Klahr, Director of Analytics Product Management at Snowflake, pointing to the initiative’s potential to simplify data operations and “prepare organisations to build the next generation of AI applications.”

As a member, Denodo will contribute to the community-driven standard for semantic model sharing — a capability it says is already embedded in its platform, which provides live access to operational and analytical data across hybrid, multi-cloud, and sovereign environments.

What this means for the market

The OSI’s emergence reflects a broader inflection point in enterprise data infrastructure. As organisations have layered AI capabilities onto existing data stacks, a critical structural problem has come into sharper focus: the same business metric — revenue, churn, active users — can be defined differently depending on which tool, team, or cloud environment is doing the calculating. That inconsistency has become an active barrier to deploying AI with any degree of reliability.

What OSI is attempting to do is address that at the specification level — before it becomes an integration problem for individual vendors to solve in isolation. That is a meaningful shift. Historically, semantic layer fragmentation has been treated as a product differentiator: each vendor maintained proprietary definitions and metrics as a form of lock-in. A move towards an open, shared standard signals that the industry may be concluding that fragmentation is now more damaging to collective AI adoption than it is beneficial to any single commercial interest.

For buyers, the practical implication is meaningful. Enterprises running multi-cloud or hybrid environments — the overwhelming majority of large organisations, have long struggled with metrics that diverge between platforms. An open specification would, in theory, allow a data product built in one environment to carry its business context into another without manual reconciliation. That reduces both integration cost and the risk of AI models being trained or queried against subtly inconsistent definitions.

Denodo’s participation is strategically coherent. The company has built its proposition around a logical, vendor-neutral semantic layer that sits above disparate data sources. Joining an initiative that legitimises and standardises that architectural concept reinforces its market position, particularly as competition in the semantic layer space has intensified, with players including dbt Labs, AtScale, and Cube vying for the same enterprise mindshare.

The broader question is whether OSI achieves the breadth of adoption required to function as a genuine standard rather than a Snowflake-adjacent specification. Snowflake’s leadership role is both an asset and a complication: it brings institutional weight and distribution, but may give pause to vendors for whom Snowflake is a competitor. How OSI manages that tension — and which major platform vendors remain absent from its membership — will determine how far its influence extends.