UnifiedModel (UModel) is a vendor-neutral semantic runtime designed to make enterprise data understandable to AI agents, facilitating data governance and operational intelligence. It transforms fragmented data schemas, entities, and relationships into a unified, workspace-scoped graph context, enabling humans and AI to query and reason over complex enterprise information consistently. This accelerates AI initiatives, reduces data governance costs, and prevents vendor lock-in by establishing a live, programmable semantic operating system.

How It Works

UModel operates as a local service, constructing a workspace-scoped object graph from model packs that define enterprise semantics. It ingests runtime entities and topology relations, exposing a unified read surface via SPL (.umodel, .entity, .topo) and REST APIs. This approach provides a consistent semantic standard across diverse platforms and tools, enabling AI agents to navigate and utilize enterprise data contextually. Its architecture prioritizes vendor neutrality and a programmable runtime, moving beyond static data dictionaries towards a dynamic semantic layer for AI collaboration.

Quick Start & Requirements

• Primary Install/Run: Execute make quickstart for a local API and Web UI with a preloaded demo workspace.
• Prerequisites: Go 1.22+, Make, Node.js 22+, pnpm 9+ (or npm exec). Verify with make check-env.
• Resource Footprint: GRAPHSTORE=memory is used for quickstart demos (ephemeral). Persistent local storage uses file.memory with --data.
• Links: Web UI: http://localhost:5173; Documentation: docs/README.md.

Highlighted Details

• Agent Skills: Loadable skills enable AI agents to directly interact with UModel for data retrieval and model-guided root-cause analysis.
• GraphStore Providers: Supports memory, file.memory (persistent JSON), and local.ladybug backends.
• SDKs: Auto-generated Go and Python SDKs are available for client integration.
• Unified Query Surface: Access models, entities, and topology via SPL, REST, CLI, and SDK using public contracts.

Maintenance & Community

• Contribution: Guidelines available in CONTRIBUTING.md.
• Support: Information in SUPPORT.md.
• Code of Conduct: CODE_OF_CONDUCT.md is enforced.
• Changelog: CHANGELOG.md tracks updates.
• No explicit community channels (e.g., Slack, Discord) are detailed in the provided text.

Licensing & Compatibility

• License: Apache-2.0.
• Compatibility: The Apache-2.0 license permits commercial use and integration with closed-source projects.

Limitations & Caveats

The open-source core focuses on local operation, public contracts, semantic modeling, and agent integration. Cloud-hosted control planes, multi-tenant authorization, and certain domain-specific APIs outside the core Query Service are explicitly excluded from this repository.