Skip to main content

​
Set Tags to ML Model Workflow

​
Overview

The Set Tags to ML Model Workflow automatically evaluates machine learning model metadata and assigns the appropriate governance tag based on whether key documentation is complete. This ensures that every ML model in the catalog is clearly marked as Complete or Incomplete, improving transparency, governance, and readiness for downstream use.

​
How the Workflow Works

  1. Workflow Start The workflow is triggered for an ML model entity. It begins by reading the model’s metadata.
  2. Check: NON-NULL Description The workflow evaluates whether the ML model contains a valid, non-empty Description field. This field is required to consider the model sufficiently documented.
  3. Decision Outcome
    • If the Description is present, the workflow follows the TRUE branch.
    • If the Description is missing or empty, it follows the FALSE branch.
  4. TRUE Path: Set Complete Tag When the model passes the description check:
    • The workflow automatically assigns the Complete tag to the ML model.
    • The workflow then proceeds to the End step.
  5. FALSE Path: Set Incomplete Tag If the model does not have a valid description:
    • The workflow assigns the Incomplete tag.
    • The workflow then ends.
  6. Workflow Completion After the appropriate tag is assigned, the workflow concludes. No human intervention is required.

​
Workflow Diagram

Below is the visual representation of the Set Tags to ML Model Workflow: Set Tags To MLModel Workflow

​
Before Workflow Execution

Description is set for the Machine Learning Model and Tags are Empty Set Tags To MLModel Workflow Description is not set and empty for the Machine Learning Model and Tags are Empty Set Tags To MLModel Workflow

​
After Workflow Execution

Complete Tag is populated Set Tags To MLModel Workflow InComplete Tag is populated Set Tags To MLModel Workflow

​
Key Features

  • Automated documentation validation Ensures ML models meet basic metadata requirements.
  • Automatic tag assignment Tags the model as Complete or Incomplete without manual action.
  • Clear governance signals Stakeholders can quickly determine whether a model has sufficient documentation.
  • Lightweight and fully automated No approval or human-in-the-loop required.

​
Example Outcomes

ConditionAssigned Tag
ML model contains a non-null DescriptionComplete
ML model missing/empty DescriptionIncomplete

​
Why This Workflow Matters

  • Ensures minimum documentation quality for ML models
  • Helps data scientists and consumers quickly identify model readiness
  • Improves governance and catalog hygiene
  • Reduces manual tagging work
  • Enables filtering and reporting based on documentation completeness

​
Summary

The Set Tags to ML Model Workflow provides a simple yet powerful automation that keeps ML model metadata aligned with governance standards. By validating the presence of a model description and assigning the correct tag based on completeness, the workflow enhances transparency, supports quality assurance, and streamlines metadata management across the ML ecosystem.