Versioning and deploying object-centric data models
Object-centric data model (OCDM) versioning and deployment provides a governance framework to capture asset snapshots, track configuration changes, and safely transport data models through your development pipeline.
This framework establishes a rigorous change-management workflow for data engineers. Isolating drafts from active production environments allows modifications to objects, events, and transformations, such as adding custom vendor attributes to an Accounts Payable process, without disrupting live business cockpits or downstream production analytics.
For a video overview of the versioning and deployment features:
OCDM versioning and deployment lifecycle
An object-centric data model moves through three distinct lifecycle states. Understanding these states ensures consistent governance across your data ecosystem:
Model state | Operational definition | Visibility and impact |
|---|---|---|
Draft | The active working environment where engineers modify objects, attributes, and event configurations. | Visible only to data developers; has no impact on active data pool executions or end-user dashboards. |
Version | An immutable, timestamped, and semantically numbered snapshot of the draft assets at a specific point in time. | Stored in the Package History repository; functions as a permanent restore point or deployable asset. |
Deployed version | A specific version activated within a designated target data pool to execute live data transformations. | Powers downstream analytics. Depending on the target, it drives either isolated validation or production reporting. |
Environment segmentation
The deployment engine separates development testing from live production execution to maintain system stability:
Development environment: Deploying to development generates data jobs with the
test:ocpm-data-jobprefix. Run these jobs to validate custom transformation logic and test schema changes against sample datasets.Production environment: Deploying to production generates data jobs with the standard
ocpm-data-jobprefix. Run these jobs to process live enterprise data and refresh customer-facing process mining applications.