Data Product Catalog

The Data Product Catalog application in Dawiso helps organizations document, manage, and share data products across departments. It supports decentralized data strategies such as Data Mesh or Data Fabric.

General overview

In traditional data governance, data is often stored and managed in a centralized data warehouse, with standardized processes and reporting handled by a central team. But modern, agile organizations (especially at scale) cannot always rely on a central data warehouse model. Teams need autonomy and speed, and often manage their own data infrastructure.

That is where the concept of data products comes in.

Each team, from marketing and product to analytics and engineering, may produce and consume their own data products: datasets, APIs, reports, or streaming outputs. These products are shared across the organization and reused by other teams. However, without structure, visibility, or ownership, this decentralized model can lead to chaos.

The Data Product Catalog provides a way to map data products. Every team can see what exists, who owns it, and how it can be used.

Data products can take many forms depending on your organization’s needs and technology stack. For example:

• A Snowflake view combining sales and customer data, shared across multiple departments.
• A Kafka stream delivering real-time clickstream data to product and marketing teams.
• A Power BI dashboard used for operational reporting, with documented input sources and owner.
• A REST API exposing key reference data to external systems.
• A Google BigQuery dataset sourced from marketing platforms for campaign analytics.
• A Python-generated file feed regularly exported and consumed by another team or system.

Data Market

Once your data products are fully documented and published, they become available in the Data Market, which allows all users to:

• Discover available data products across spaces
• Search for specific platforms or technologies (e.g., “all products using Google Cloud Platform”)
• Access and request data products in a structured, governed way

Data Market supports data discoverability and democratization, making it easy for anyone in the organization to find and reuse trusted data products.

Each card lists the product’s output ports by name, and shows the PII and confidentiality classification calculated for each port when the product was published. A consumer can therefore judge how sensitive a product is before requesting access, rather than after. A port whose assets carry no classification shows N/A.

To create a product that appears here, see Create a data product.

Each data product has a visibility classification that determines who can request access:

Visibility	Who can request
Public	Any user
Dataspace	Users within the same dataspace
Organisation	Users within the organisation
Invitation	Users who have received an explicit invitation
Private	Access is restricted; contact the data owner

Request access to a data product

1. Browse the Data Market.
2. Locate the data product(s) you need access to and click Add to cart.
   Tip

   Data Market shows which products you already have access to, or for which you requested access. See the table below for what each request status means.

3. Open the cart to review your selection. It lists each product’s owner, sensitivity classification, visibility, and the approvers who will review the request.
4. Click Checkout. What happens next depends on your organization’s checkout target:
   • Gate (the default): a dedicated Checkout page opens instead of a dialog, with one card per product covering its ports, personal data access, and row scope, plus a live preview of what Gate would grant. See If checkout uses Gate.
   • Jira Service Management (JSM) or ServiceNow: a dialog opens. Fill in:
     • Reason: Explain why you need access to the product.
     • Duration: Set how long you need the product for. Some products cap the maximum based on their visibility level.
     • Use case: Provide the use case for which you need the product, e.g., ad hoc analysis, recurring report, ML training.
     • Row scope: Optionally, select your row scope.
5. Submit the request:
   • Gate: click Submit all. The button stays disabled while any card is blocked. Submission reports each product’s outcome right away: access can be granted immediately, wait for an approver, or be blocked, and one product can partly succeed while the rest of the cart goes through. See If checkout uses Gate for what each outcome means.
   • JSM or ServiceNow: submit the dialog to raise the request in that system. The product’s designated approvers review it there.

For a Gate request, track it under My Access, where its status is one of the following:

Status	Meaning
Pending	Waiting for review.
Waiting for fulfillment	Approved and provisioning has started. An external system now needs to complete the grant.
Provisioned	Access is active and can be used.
Expiring soon	Access is still active but nearing its end date.
Rejected	Approver denied the request.
Cancelled	You (or an admin) cancelled the request before it was fulfilled.
Failed	Provisioning failed but may be retried.
Permanently failed	Provisioning failed and won’t be retried automatically. Contact the data owner for assistance.

JSM and ServiceNow requests aren’t tracked in Dawiso: neither system reports status back, so My Access doesn’t list them. Check the request in JSM or ServiceNow directly.

Data Product Catalog object types

The Data Product Catalog is organized around three core object types: Domains, Data Products, and Contracts.

Object type	Description
Product Domain	Domains represent logical areas of the organization, such as Sales, Marketing, or Customer. They serve as containers for related data products.
Data Product	A data product is an object that represents a structured, reusable set of data assets designed to deliver value to internal or external consumers.
Contract	Contracts document how data flows between data products and under what terms it can be used.

Data Product Catalog relations

There are two types of relations for Data Product Catalog objects:

Contains

This relation connects Data Products and other objects in Dawiso, such as datasets or business terms.

• Contains: The data product contains information from another Dawiso object.
• Is contained in: The object is used in a data product.

Has input/output

A data product can be linked to other products via contracts. There are two types of relations for data products:

• Has input: The data product is a data consumer, since it uses other data products as its input.
• Outputs contract: The data product is a data producer, since it is used as input in other data products.

Key features

Contracts: Document usage and responsibilities

To support these relations, Dawiso allows the documentation of contracts between data products. A contract formalizes the connection between the data producer (the source product) and the data consumer (the dependent product). It specifies how the data can be used and under what conditions.

Each contract:

• Connects two data products with a clear producer–consumer relationship.
• Can include SLA details, such as bug resolution times or update frequency.
• Allows for issue reporting and communication between product owners.
• Appears directly in the lineage view, making data dependencies and responsibilities easy to trace.

Together, lineage and contracts provide a complete picture of how and under what conditions data products interact.

Data lineage: Understand how your data products connect

One of the main features of the Data Product Catalog is the data product lineage. It visually maps the relations between data products, showing which products are built from others, and which products depend on them.

For example, a team might combine reports from Marketing, Sales, and Customer Support into a new product. In Dawiso, this relation is clearly shown in the lineage view, helping everyone understand where data comes from and how it flows across teams.

This improves transparency, with impact analysis, ownership clarity, and supports collaboration between data producers and consumers.

Workflow logic

Each data product in Dawiso includes key metadata and supports a workflow that mirrors its lifecycle, from idea to delivery.

Design → In Development → Ready for Publishing → Published

This lifecycle workflow supports both the documentation of existing products and the creation of new ones.

Add a Data Product Catalog to your space

To start using the Data Product Catalog application, first add it to your space.

1. In the top-right corner of your space, click the + icon and select Add application.
2. Select Data Product Catalog and the Data Product Workflow.

Add a domain

Start by adding a domain, which serves as a container for categorizing your data products.

1. In your Data Product Catalog app, click Create in the top navigation bar.
2. Add an object with the Data Domain object type.

Create a data product

The space must have the Data Product Catalog application assigned before you can create a product there. If your space isn’t listed, ask an administrator to assign it.

1. While on the domain’s page, click Create in the top navigation bar. Create is context-aware, so this preselects the product’s space and domain. Name the product and add a description and value proposition. Both appear on the product’s Data Market card.
2. Optionally connect the product to business terms, use cases, related data products, and labels, so consumers can judge relevance before opening it.
3. Assign owners, stewards, custodians, subject-matter experts, and producers so consumers know who to contact.
4. Pick a technology to unlock the asset list, then narrow it by asset type. You can select more than one technology. Each asset shows its PII and Confidentiality classification.
5. Group the assets into output ports and record the input ports the product consumes. Each port shows an aggregated PII/Confidentiality badge: the highest classification among its assets wins, and the asset that set it is marked in the port’s table. A port with no classified assets shows N/A, meaning unclassified, not confirmed free of PII.
6. Attach a data contract to each output port, either a new one or an existing one from the product’s domain.
7. Review the product, then Save for later or Publish now.

Propose a data product with Agent Studio

Some object types can be flagged as candidates for a data product. On a flagged object, Dawiso can ask Agent Studio to draft a data product proposal. You then start from a pre-filled data product wizard instead of documenting the product from scratch. By default, the flag is available on the Databricks table and Databricks view object types.

1. Open the flagged object.
2. In the object header, click Promote to data product.

You will receive a notification as soon as the process finishes.

Outcome	What happens
Ready	Your data product proposal is ready to Open in wizard, where the data product the proposal is filled in.
Refused	The proposal was not created. See the notification for the reason, and any warnings provided by the agent.
Failed	The data product proposal failed to be prepared.
Timed out	Timeout occurs for processes that haven’t finished after roughly 15 minutes.

A proposal is refused for one of these reasons:

• The object can’t become a data product.
• Data products aren’t configured for this environment.
• The object is already part of a data product.
• No single domain could be determined.
• The object has too many dependencies.
• The proposal could not be saved.
• An internal error occurred.

Export data products

You can export Data Products as an OSI Semantic Model YAML file for use in BI, AI, and semantic-modeling tools such as Snowflake. The export describes the Output Ports of the selected Data Products, including tables, columns, and foreign-key relationships, in the Open Semantic Interchange (OSI) format.

For step-by-step instructions, see Export a data product as OSI YAML.