Data product release version management

Blindata provides release version management for data products, enabling organizations to publish, govern, and trace immutable versions of their data product definitions across the lifecycle. Each release is identified by a semantic version (e.g. 1.0.1) and linked to a Git tag, ensuring full traceability from source control to deployment.

Release publishing implements shift-left governance: contracts, metadata, and policies are validated before a version is exposed to consumers. Once a version is published, it can be promoted across lifecycle stages (e.g. dev → test → prod); promotion and automation are managed by the Data Product DevOps Management module.

Key capabilities

Use cases

• Governed releases: Tag a descriptor in Git, generate changelog from commits, publish after policy validation.
• Release history and rollback: Browse all versions with changelogs and lifecycle status; identify what changed and when.

The Builder is the central interface for managing releases: view the latest version, browse all releases with changelogs, and create new ones via Git tags or file upload.

The Builder page: release card and DevOps lifecycle

From the Builder page of your data product you get a clear overview of the release state and DevOps lifecycle:

Release versions list and changelog

Click ALL RELEASES to open the release versions list:

Each version is shown as a card with:

• Version number (e.g. 1.0.1) and publication date
• Publisher (user who created the release)
• Policies Validation status (Approved when checks passed)
• Changes (changelog): commits included in that release, with message, short hash, and author

You can search releases and use + CREATE NEW RELEASE to add a new version. Click a version number to open its detail page.

Version detail: documentation and DevOps activities

From the list, click a version to open its detail page:

• DOCUMENTATION tab: changelog (commits in this release), release notes, and metadata.
• DATAOPS tab: lifecycle activities for this version across stages.

Under DATAOPS you see:

• A timeline for the version’s stages (e.g. dev → prod)
• The Version Activities table: stage, status (PROCESSED, ABORTED, etc.), creation time, start and end times
• PLAN ACTIVITY to start new lifecycle actions for this version

Use the edit or delete icons when you need to update or remove this version.

Create a new release version

You can create a new release in two ways: by Git tag (recommended) or by uploading a descriptor file (for legacy workflows).

Option 1: Create from a Git tag (recommended)

1. On the Builder page, click + CREATE NEW RELEASE, or go to the releases list and click + CREATE NEW RELEASE.
2. In Create new release version, choose a tag:
   • Version tag: Pick an existing tag (e.g. 1.0.1) or create a new one.
   • Previous tag: Select the base tag for comparison (e.g. 1.0.0). This is computed automatically as the latest release before the selected version tag, but you can change it to compare against a different tag.
3. Click REFRESH RELEASE NOTE to generate the release notes. The changelog is built by reading the commits between the previous tag and the version tag—each commit message is listed in the Changes section. You can then edit the content in the rich text editor if needed.
4. Edit the Title if needed (it is pre-filled from the version tag).
5. Optionally adjust the Changes content in the rich text editor.
6. Click PUBLISH.

The changelog is derived from the commit messages in the selected range, so keep commits clear and consistent (e.g. conventional commits).

Option 2: Create by uploading a descriptor file (legacy)

For data products without a connected Git repository, or when you need to publish from a standalone descriptor, use the file upload flow:

1. On the Create new release version page, click the options menu (⋮) in the top right corner and select Upload descriptor file.
2. Upload a DPDS file in JSON or YAML format:
   • Drag and drop the file into the upload area, or click to select it.
3. Check the Preview to ensure the descriptor is correct.
4. Click UPLOAD to create the release.

The descriptor must follow the Data Product Descriptor Specification (DPDS) . This option is intended for legacy workflows; for new data products, prefer Git-based releases for better traceability.

Summary