Apache Ossie: Microsoft, Google Back Shared Semantic Model Standard

What happened

Microsoft and Google have joined Apache Ossie, an open project to define a shared specification for exchanging semantic models across data, analytics and AI platforms. More than 60 companies already back the project, including Databricks, Informatica, Mistral AI, Nvidia, Oracle, Salesforce and Snowflake, according to InfoWorld.

The project started as Open Semantic Interchange (OSI). It was renamed Apache Ossie after it entered the Apache Incubator in June. Semantic models, which describe datasets, fields, relationships, metrics and AI context, are represented in JSON and YAML so they can move between platforms such as Snowflake, Databricks, Tableau, ThoughtSpot and Sigma.

The design uses a hub-and-spoke model. Ossie serves as the common semantic format at the centre, while converters translate between it and each vendor's own format. Analysts cited by InfoWorld said the backing makes interoperable analytics platforms more likely in future, but does not guarantee that vendor lock-in will fall away.

Why it matters

For organisations running several data and AI tools, the practical problem is rarely storage. It is that the same business term, such as "active customer" or "net revenue", can be defined differently in each system. A shared semantic format targets that gap directly, so that a metric defined once can be used consistently by dashboards, analysts and AI assistants without being rebuilt for each platform.

The significance lies in the operating model as much as the technology. If definitions become portable, leaders gain more room to change or combine platforms during digital transformation programmes without rewriting the logic that frontline teams and customers depend on.

By the numbers

• 60+ companies have backed the Apache Ossie project, including Microsoft and Google's recent support.
• June is when the project, then named Open Semantic Interchange, was accepted into the Apache Incubator as Apache Ossie.

The Renascence take

Most coverage will treat this as a vendor-alliance story. The more useful question is whether a definition of "customer" survives the journey between platforms with its meaning intact.

Interoperability tested only by whether files load is a shallow test. Customers and staff experience the consequences when one team's metric disagrees with another's, and trust in data erodes quickly. Operators should audit where their core metric definitions actually live, write portability of those definitions into vendor contracts, and pilot one converter before committing to a platform change.

--- **Meta title:** Apache Ossie: Microsoft and Google Back Open Semantic Models (59 chars) **Meta description:** Microsoft and Google join Apache Ossie, an open spec for semantic models across data and AI platforms. 60+ firms back it; analysts say lock-in may persist. (150 chars) **Title:** Apache Ossie: Microsoft and Google Back Interoperable AI Data **Target keywords:** Apache Ossie; semantic models; data interoperability; enterprise AI platforms; vendor lock-in; Open Semantic Interchange; Microsoft Google AI; AI data platforms