Reports Layer
Dashboards, visuals, and pages are represented as report objects with reusable metadata and layout semantics.
The first open standard for unified BI assets.
OpenBI gives organizations one shared model to describe and manage BI assets across platforms, built for business teams and interpretable by AI agents.
Bring your AI agent of choice. OpenBI MCP Server makes it work on your BI projects, your platforms, your way.
BI teams are still blocked by fragmentation
Business Intelligence delivery is spread across disconnected platforms and incompatible asset formats. Teams duplicate effort, repeat manual steps, and struggle to scale automation.
- Power BI, Qlik, SAP BO, Tableau each speak a different language.
- Reports, models, and dataflows are managed in tool-specific silos.
- AI potential is constrained when systems cannot share the same structure.
OpenBI is the standard layer across BI platforms
OpenBI defines a common representation for frontend and backend BI artifacts, including reports, semantic models, dataflows, paginated reports, apps, and more.
With a shared semantic model, organizations can build once, reason consistently, and execute across different BI ecosystems.
How the OpenBI model is composed
OpenBI models BI assets as connected components in one interoperable graph. Reports, semantic models, databases, and data sources are represented as standardized objects and linked through explicit relationships.
Semantic Models Layer
Tables, measures, and relationships are mapped in a unified structure so business logic remains portable and inspectable.
Databases Layer
Physical storage structures are linked to semantic entities, creating a clear path from data storage to business meaning.
Data Sources Layer
Files, APIs, and upstream systems are normalized as data source objects, enabling traceable lineage across platforms.
AI-ready by design, not by workaround
OpenBI is designed to be interpreted and produced by AI systems. OpenBI MCP Server is the bridge that allows agents to operate on BI assets through simple, extensible interfaces.
Bring Your Own Agent
Use Claude, Codex, Gemini, OpenCode, or any compatible agent through one MCP interface.
Customize to Your BI Landscape
Adapt agent behavior to your BI assets, platform specifics, and internal delivery patterns without rebuilding everything.
Platform-Ready System Prompts
Start faster with ready-to-use prompts for common BI platforms, then extend them for your standards.
Enable BI as Code
Automate recurring BI work and orchestrate high-level BI workflows with a shared model at the core.
Open project, practical foundation
OpenBI is built as an open foundation for the BI community and AI developers. The ecosystem includes:
- OpenBI Models for standardized asset representation.
- OpenBI.MCP.Server to connect AI agents to BI projects.
The goal is simple: reduce manual BI work, increase interoperability, and unlock reliable AI-driven execution.
One standard for BI. Any platform. Any AI agent.
Build and manage BI assets with open models and a customizable MCP execution layer.