Basics

Open Data Product Standard (ODPS)

Dr. Simon Harrer

Co-Founder & CEO, Entropy Data · 2026-09-15

The Open Data Product Standard (ODPS) is an open YAML specification for describing a data product: what it is, which data it consumes and provides, who owns it, and how to use it. It is governed by Bitol, a Linux Foundation AI & Data project, and licensed under Apache 2.0.

ODPS is the sister of the Open Data Contract Standard (ODCS). ODPS describes the product, ODCS describes the data behind each of its ports. The current version is v1.1.0.

Example

apiVersion: v1.1.0
kind: DataProduct

id: fbe8d147-28db-4f1d-bedf-a3fe9f458427
name: Customer
version: v1.0.0
type: aggregate
domain: sales
status: active

description:
  purpose: Enterprise view of a customer.
  usage: Use consolidated_transactions for reporting.

inputPorts:
  - name: payments
    version: 2.0.0
    contractId: dbb7b1eb-7628-436e-8914-2a00638ba6da

outputPorts:
  - name: consolidated_transactions
    description: Joined and aggregated transactions
    type: tables
    version: 1.0.0
    contractId: a44978be-1fe0-4226-b840-1b715bc25c63

context:
  instructions: Refresh latency is 4 hours.
  constraints:
    - constraint: Do not expose individual customer PII.

team:
  name: customer-team
  members:
    - username: john.doe@example.com
      role: owner

support:
  - channel: "#customer-data"
    tool: slack
    url: https://example.slack.com/archives/C1234567890

Content

These are the building blocks of an ODPS data product:

• Fundamentals: id, name, version, domain, status, description, tags, and since v1.1.0 a type such as sourceAligned, aggregate, or consumerAligned.
• Input ports: the data the product consumes, each optionally pointing to an ODCS data contract.
• Output ports: the data the product provides, each with its data contract, input dependencies, and an optional SBOM. At least one is required.
• Management ports: REST endpoints or topics for discoverability, observability, and control.
• Context: instructions, verified statements, and constraints that tell AI agents how to use the data product and its output ports.
• Team and support: owners and members, plus the channels where consumers get help.
• Custom properties and authoritative definitions: extensions and links to external documentation, shared with ODCS.

v1.1.0 also adds synonyms, deprecation flags, stable ids on every array object, and ${VAR_NAME} variables. See the release notes for details.

Tooling

Entropy Data uses ODPS to build its data product marketplace, and supports v1.1.0 natively. The open-source Data Product CLI creates, lints, and publishes ODPS files:

pip install dataproduct-cli
dataproduct init customer.odps.yaml
dataproduct lint customer.odps.yaml
dataproduct publish customer.odps.yaml

Context on an ODPS v1.1 data product in Entropy Data, for consumers and AI agents

Sign up now for free, or explore the clickable demo of Entropy Data.

About the author

Dr. Simon Harrer

LinkedIn of Simon Harrer

Co-Founder & CEO, Entropy Data

Simon builds tools that help data, people, and AI work together. He holds a PhD in distributed systems, spent seven years at INNOQ as a software architect and consultant, and co-founded Entropy Data, where he leads the company as CEO.

Get in touch with Simon →