The Universal Standard for Semantic Model Exchange

An open semantic model specification enabling semantic metadata interchange across analytics, AI, and BI platforms, providing a vendor-neutral single source of truth for your semantic data.

View Specification Get Involved

About OSI

The Open Semantic Interchange (OSI) is a collaborative, open-source effort dedicated to standardizing and streamlining semantic model definitions across the data analytics, AI, and BI ecosystem.

🔄

Interoperability

Seamlessly exchange semantic models between AI agents, BI platforms, and analytics tools.

✅

Consistency

Maintain consistent data defi nitions and values across every platform in your ecosystem.

🌐

Vendor-Agnostic

A common standard that works across all vendors, eliminating tool-specific inconsistencies and lock-in.

🚀

Efficiency

Reduce engineering debt and accelerate innovation by using a unified, governed semantic foundation.

Why OSI?

The Challenge (Semantic Fragmentation)

Metric Drift: Inconsistent KPIs across different dashboards.
Manual Translation: Costly, error-prone reconciliation efforts.
Hallucinations: Unreliable AI grounding from conflicting data logic.
Integration Debt: Complex N-to-N custom integrations between proprietary tools.

The Solution

Single Source of Truth: Unified semantic and metric definitions.
Native Interoperability: Direct exchange between platforms and AI agents.
Trusted AI Grounding: Agents reasoning accurately based on business logic.
Reduced TCO: Lower costs through automated model exchange.

Working Group Members

Leading organizations collaborating to build the future of semantic interchange

Alation

Atlan

AtScale

BlackRock

Blue Yonder

Coalesce

Collate

Collibra

Credible

Cube

Databricks

DataHub

dbt Labs

Domo

Elementum AI

Firebolt

Hex

Honeydew

Informatica

JetBrains

Lightdash

Mistral AI

Omni

Preset

Qlik

RelationalAI

Salesforce

Select Star

Sigma

Snowflake

Starburst Data

Strategy

ThoughtSpot

Specification

Core Classes

Semantic Model: The top-level container that represents a complete semantic model, including datasets, relationships, and metrics.
Data Sets: Logical datasets represent business entities or concepts (fact and dimension tables). They contain fields and define the structure of the data
Fields: row-level attributes that can be used for grouping, filtering, and in metric expressions.
Measures: Quantitative measures defined on business data, representing key calculations like sums, averages, ratios, etc. Metrics are defined at the semantic model level and can span multiple datasets.
Dimensions: Categorical attributes (Where, When, Who).
Relationships: Relationships define how logical datasets are connected through foreign key constraints. They support both simple and composite keys.

Current Working Groups

Advanced Metrics & Expression Language
Composability
Catalog Integration
Ontology representation
Model converters and developer tools

View OSI Repository

Latest Updates

Announcement

Open Semantic Interchange: A New AI Standard

Introducing OSI, a collaborative initiative to create a vendor-agnostic standard for semantic model exchange across AI and BI platforms.

Expansion

OSI Initiative Grows with New Partners

Major industry leaders, including JPMC and Collibra, have joined the working group to advance semantic data standards

How to Contribute?

Join our open-source community on GitHub. Decouple your semantic logic from proprietary platforms and help shape the future.

Review the OSI repository and documentation for the current spec.

Have a use case or an idea for the spec evolution that you would like to discuss with the community? Start a discussion here.

Have a suggestion to modify the existing spec? Raise a Pull Request in the repo.

Have tools or converters you would like to add? Raise a Pull Request in the repo.

Join the Working Group

Whether you are a vendor or customer, ensure your voice is heard. Register your interest to influence the evolution of the OSI specification.