apache/ossie

Apache Ossie, industry wide specification effort to standardize how we exchange semantic metadata across analytics, AI and BI platforms, providing a vendor neutral, single source of truth for semantic data

Apache Ossie (incubating)

What it is

  • An Apache‑incubating open‑source project that defines a vendor‑agnostic semantic model specification for data‑analytics, AI, and business‑intelligence tools.
  • The spec is expressed in JSON and YAML, with a machine‑readable schema, so any tool can import or export the same business‑logic definitions (KPIs, metrics, dimensions, etc.).

Why it matters

  • Today many BI/AI platforms each maintain their own definition of the same KPI, leading to duplicated effort and inconsistent AI‑generated insights.
  • By providing a single “source of truth” that AI agents, BI dashboards, and data pipelines can all read, Ossie aims to eliminate that semantic fragmentation and make AI outputs more reliable.

Key parts of the repository

Directory Content
core-spec/ The official Ossie specification (spec.md), the JSON schema (ossie-schema.json) and YAML schema (spec.yaml).
examples/ Sample semantic models, including a full TPC‑DS benchmark model, showing how real‑world metrics are expressed.
validation/ Small utilities that check a model against the schema, useful for developers building converters or exporters.
docs/ Human‑readable documentation, overview, and usage guides.

How to use it

  1. Read the spec – start with core-spec/spec.md to understand the data structures (entities, attributes, relationships, KPI definitions).
  2. Create a model – write a JSON or YAML file following the schema; you can copy one from examples/ and adapt it.
  3. Validate – run the validation tools (validation/) to ensure your file conforms to ossie-schema.json.
  4. Integrate – build or use a converter (see the companion apache/ossie-converters repo) to translate between Ossie and the format your BI/AI tool expects (e.g., dbt, GoodData, Salesforce).

Getting involved

  • Contribute: see CONTRIBUTING.md for the process to propose spec changes or add code.
  • Roadmap: ROADMAP.md lists current working groups and upcoming features.
  • Community: join the GitHub Discussions, file issues, or chat on the project Slack (link in the README).

Bottom line Apache Ossie is not a library that runs AI models; it is a standard that lets AI agents and analytics platforms speak the same language about business concepts, reducing mismatches and making AI‑driven insights more trustworthy.

Related

  • Project
    agentic-commerce-protocol/agentic-commerce-protocol

    An open standard and interaction model maintained by OpenAI and Stripe that enables AI agents to seamlessly conduct commercial transactions between buyers and businesses.

  • Project
    Canner/WrenAI

    An open-source generative BI engine that provides a governed semantic layer for AI agents to turn business questions into correct SQL and deployable dashboards.

  • Project
    agntcy/oasf

    OASF is an open‑source, versioned JSON schema for describing AI agents—its capabilities, skills, and domains. It ships with a validation server, hot‑reload tooling, and LLM‑assisted record generation via the MCP server, enabling interoperable agent catalogs and discovery.

  • Project
    roboflow/supervision

    A computer vision toolkit that provides model-agnostic building blocks for data loading, visualization, and dataset management to accelerate the development of vision applications.