AI Workflows can be paused and edited at any time.

This allows teams to:

Workflows are designed to be flexible and continuously adjustable.

Pausing a workflow temporarily stops automation activity.

This is useful when:

Paused workflows can later be resumed.

When a workflow is paused:

Existing results and workflow configurations remain unchanged.

To pause a workflow:

The workflow status will update to Paused.

Paused workflows can be resumed at any time.

To resume a workflow:

The workflow will continue processing products according to the configured trigger rules.

For continuous workflows:

Workflows can be edited after creation.

You can modify:

This makes workflows adaptable as catalog requirements evolve.

Trigger rules determine which products enter the workflow.

You may want to edit rules when:

Example:
Original rule:

Updated rule:

This creates more targeted processing.

Workflow actions can also be updated.

Examples:

This allows continuous optimization of AI output quality.

Prompts are one of the most commonly adjusted workflow settings.

You may update prompts to:

Prompt optimization is a normal part of workflow management.

Moderation settings can be updated to improve quality control.

Examples:

Different workflows may require different moderation strategies.

Workflows can sometimes be edited while active, but it is usually recommended to pause large workflows before making major changes.

This helps avoid:

For major updates:

After making workflow changes, it is strongly recommended to:

before restarting large executions.

If workflows process too many products:

This improves output relevance and moderation efficiency.

Weak prompts often create:

Refined prompts usually improve:

Very large workflows may become difficult to manage.

Instead of one massive workflow, create focused workflows such as:

This improves maintainability and troubleshooting.

Common reasons for editing workflows include:

Workflow adjustments help improve long term automation quality.

Example:
A webshop notices that a workflow enriches too many unrelated products.

Original trigger:

Updated trigger:

The workflow now targets only relevant products, improving enrichment quality and reducing unnecessary AI processing.

Good naming improves maintenance.

Examples:

Always validate:

before large executions.

Even stable workflows should be reviewed periodically to:

For major workflow changes:

This reduces operational risks.

Example:
A webshop updates its SEO strategy.

Workflow changes:

This allows workflows to evolve with changing business requirements.