Data Products

A Data Product packages curated data assets under a single entity with defined ownership, type, and visibility. Create one from the Data Marketplace, or from within a Domain if it belongs to a specific one. This guide walks through both paths, plus adding assets, input and output ports, and finding a data product’s assets from the Explore page.

Add a Data Product

To create a new data product, follow these steps:

Step 1: Access Data Products

1. In the left navigation bar, click Data Marketplace, then select Data Products.
2. Click Add Data Product.

Step 2: Enter Data Product Details

• Icon and Color: Select an icon and color to visually identify the data product in the catalog.
• Name (required): Enter a unique identifier for the data product.
• Display Name: Enter a human-readable label shown throughout the UI.
• Description (required): Explain the data product’s purpose, the data it contains, and any context other users need.

Step 3: Add Tags and Glossary Terms

• Tags: Select classification tags to categorize the data product.
• Glossary Terms: Link the data product to relevant glossary terms.

Step 4: Configure Data Product Settings

• Domain (required): Select the domain this data product belongs to.
• Type: Select the data product type:

  Type	Description
  Raw Data	Unprocessed source data as collected, with no transformation applied.
  Derived Data	Data produced by transforming, cleaning, or combining other data.
  Dataset	A curated collection of data packaged for reuse.
  Reports	Prepared reports that present data to business users.
  Analytic View	A modeled view of data built for analysis and reporting.
  3D Visualisation	Data presented as three-dimensional visual models.
  Algorithm	A data-driven algorithm or model offered as a product.
  Decision Support	Insights and recommendations that help people make decisions.
  Automated Decision Making	Data used to make decisions automatically, without human input.
  Data-Enhanced Product	A product that is improved by embedded data.
  Data-Driven Service	A service whose core functionality is powered by data.
  Data-Enabled Performance	Data used to improve the performance of a process or system.
  Bi-Directional	Data exchanged in both directions between provider and consumer.

• Visibility: Control who can discover this data product.

  Visibility	Access
  Private	Visible to the creator only.
  Invitation	Visible to invited users.
  Dataspace	Intended for your organization and its partner organizations in a shared data space. This label doesn’t enforce visibility.
  Organisation	Visible to all users in the organization.
  Public	Visible to anyone.

  A data space is a group of separate organizations that agree to share data with each other under common rules. The visibility values follow the Open Data Product Specification (ODPS). OpenMetadata doesn’t let you define a data space or its members. Dataspace is a label that describes the intended audience. It doesn’t change who can see the data product. Access is still controlled by roles and policies.
• Portfolio Priority: Indicate the data product’s importance in your portfolio:

  Priority	Description
  Critical	Business-critical. Many teams or key processes depend on it.
  High	Important and widely used.
  Medium	Moderately important or used by a limited set of teams.
  Low	Useful but not essential, or used rarely.

  Portfolio Priority is informational. It doesn’t affect search ranking, recommendations, or AI results. It is also used in the following ways:
  • Filter: Use the Portfolio Priority filter on the Data Products page, or filter by it through the search API.
  • MCP: The value is returned when Model Context Protocol (MCP) tools read data products.
  • ODPS: The value is included when you export or import a data product as ODPS.
  • View or change: After you create the data product, open Edit Metadata from the data product’s more actions menu to view or change the value. The data product page doesn’t show it.

Step 5: Assign People

• Owners: Select the users or teams responsible for the data product.
• Experts: Select individual users with deep knowledge of the data product.

Step 6: Save Data Product

Click Save. The data product appears in the Data Products list.

Add a Data Product from Within a Domain

If a data product belongs to a specific domain, create it directly from that domain instead:

1. Navigate to Data Marketplace > Domains, then open the domain the data product should belong to.
2. Click Add and select Data Products.
3. Follow Add a Data Product steps 2 to 6.

In the Data Products tab, you can view the list of all data products related to the selected domain.

Add Data Assets to the Data Product

1. From the Data Products list, open the data product’s detail page.
2. Click Add Assets.
3. Search for assets, or use the filters (entity type, owner, tag, tier, service, service type) to narrow the results.
4. Select the assets to include, then click Save.

By default, an asset can belong to multiple Data Products, but each Data Product must belong to the same domain as the asset itself. Admins can change this in Data Asset Rules using the Multiple Data Products are not allowed and Data Product Domain Validation rules.

Manage a Data Product

From a data product’s detail page, click the ⋮ (more actions) menu next to the version count to manage the data product itself.

• Announcements: Set up banners to inform your team of upcoming maintenance, updates, and deletions. For more information, see Announcements.
• Rename: Change the Name and Display Name of the data product.
• Style: Change the icon and badge color for the data product.
• Edit Metadata: Change the type, visibility, and portfolio priority of the data product. After you create a data product, Edit Metadata is the only place in the UI to view or change Portfolio Priority. For more information, see Edit Data Product Metadata.
• Delete: Permanently delete the data product and its metadata.
• Export as ODPS: Export the data product as an Open Data Product Specification (ODPS) v4.1 YAML file.
• Import from ODPS: Import a data product definition from an ODPS v4.1 YAML file.

Edit Data Product Metadata

To change the type, visibility, or portfolio priority of an existing data product, follow these steps:

1. Open the data product and click the ⋮ (more actions) menu next to the version count.
2. Click Edit Metadata. The Edit Data Product dialog opens.
3. Update any of the following:
   • Type: Select the data product type.
   • Visibility: Select who can discover this data product. For more information about each value, see Configure Data Product Settings.
   • Portfolio Priority: Indicate the data product’s importance: Critical, High, Medium, or Low.
   
4. Click Save.

Troubleshooting

Common issues when adding assets to Data Products:

Data Product Domain Validation Doesn’t Recognize Subdomains

Problem: An asset assigned directly to a subdomain (for example, customer_domain.finance) can’t be added to a Data Product assigned to the subdomain’s parent domain (customer_domain), even though the subdomain is nested under it. The same asset assigned directly to the parent domain can be added without issue. Cause: The Data Product Domain Validation rule currently checks for an exact domain match between an asset and a Data Product. It doesn’t yet treat a subdomain as a match for its parent domain. Status: Disabling the rule in Data Asset Rules doesn’t currently resolve this for subdomains, since the underlying domain-matching logic needs to recognize the parent-subdomain relationship independently of the toggle. This is a known limitation the OpenMetadata team is actively addressing.

Input Ports and Output Ports

Beyond general assets, a data product can declare which assets it consumes and which it produces or exposes: an ODCS/ODPS-aligned data contract concept that makes a data product’s data flow explicit for lineage-style visualization.

• Input Ports: Data assets this data product consumes. Use input ports for sources your team does not own, such as a table from another domain or an external system you read from but do not manage. An input port does not have to be one of the data product’s assets. Record an external source you only consume as an input port, and leave it off the Assets list.
• Output Ports: Data assets this data product produces or exposes to consumers. An asset must already belong to the data product before it qualifies as an output port.

Ports describe how data flows into and out of the data product, not who owns each asset. Keep the Assets list to the resources your team owns and manages, and record the external sources you consume as input ports. To add ports:

1. From the data product’s detail page, go to the Input/Output Ports tab.
2. Under Input Ports or Output Ports, click Add Port.
3. Search for and select the assets to add, then click Save.

The tab also offers a Ports Lineage view, which visualizes input and output ports as a lineage-style graph alongside the data product.

Approval for Port Changes

When a governance workflow with Approval Mode set to Enforce covers Data Products, adding or removing a port waits for approval.

• Held change: OpenMetadata does not add or remove the port until a reviewer approves the change request. Rejecting the request keeps the current ports.
• One request: A held port change and a held field change from the same person form one change request.
• Bots: Bots write port changes directly.
• Port picker: The port picker does not list table columns.
• Adding assets: The workflow of each asset gates adding that asset to a data product, not the workflow of the data product.

For more information, see Approval-Gated Changes.

View Data Assets in the Data Product from the Explore Page

1. In the left navigation bar, click Explore.
2. In the Data Assets dropdown, select Data Product.
3. All data products appear in the results. Click one to view its assets, documentation, and metadata.

Watch the video below to see data products in action.