A coalition of leading data and AI companies, including Snowflake, Salesforce, and dbt Labs, has announced the launch of Open Semantic Interchange (OSI), an open-source initiative designed to standardize how semantic layers are defined and shared across data platforms.
Addressing Fragmented Data Semantics
Enterprise AI adoption often faces challenges due to fragmented data semantics. Different analytics and machine learning tools may interpret business logic—such as metrics, hierarchies, and relationships—differently, creating inconsistencies and inefficiencies. OSI aims to solve this by providing a shared standard for portable and interoperable semantic definitions across ecosystems.
Consistent Semantics for AI and Analytics
With OSI, organizations can define business semantics once and apply them seamlessly across data warehouses, BI platforms, and AI applications. This consistency is expected to enhance data governance, accelerate AI deployment, and reduce duplication of work, helping enterprises become more AI-ready.
Community-Driven Open Standard
The initiative will remain community-driven, with Snowflake, Salesforce, and dbt Labs encouraging contributions from developers and technology providers. Early implementations are planned for integration into popular data platforms by 2026, following public feedback on technical specifications.
Shaping the Future of Scalable AI
Industry experts note that OSI could play a pivotal role in ensuring trustworthy and scalable AI by enabling machine learning systems and LLMs to operate with accurate, context-rich business logic. The consortium’s launch reflects a broader trend toward openness and collaboration as enterprises strive to unlock more value from their data in the AI era.
