After configuring and testing your workflow actions, you can execute them directly from the workflow overview.

Running workflow actions starts the AI processing pipeline for all products matching the configured trigger rules.

Depending on the workflow setup, actions may run:

Before starting a workflow action, it is strongly recommended to:

This helps prevent large scale processing errors.

To run an action:

The workflow will then begin processing matching products.

When an action starts:

Actions process products sequentially through the workflow pipeline.

Actions execute in order from left to right.

Example:

In this workflow:

Each action builds on the previous action’s output.

One-time workflows process all currently matching products once.

Example:

After execution completes, the workflow stops automatically unless manually restarted.

Continuous workflows continuously monitor the catalog.

Whenever new products match the trigger rules:

This creates ongoing automation pipelines.

While actions are running, the workflow overview displays:

This helps teams track workflow activity in real time.

Workflow actions may display statuses such as:

Statuses help identify where products currently are inside the workflow pipeline.

Each action contains a Results tab showing:

The Results tab provides detailed visibility into workflow execution.

Depending on the action configuration:

Products can pause temporarily in moderation queues before continuing through the workflow.

Certain actions display:

Reasoning explains why the AI generated specific output.

Confidence scores help identify:

This improves quality control during execution.

Approved outputs can synchronize automatically back into your connected platform.

Examples:

Depending on workflow settings:

Large workflows may take time depending on:

Products are often processed individually to maintain stability and avoid provider rate limits.

Workflows can be paused during execution.

Pausing is useful when:

Paused workflows can later be resumed.

Useful for:

Typically uses one-time workflows.

Useful for:

Typically uses continuous workflows.

Before processing large catalogs:

This reduces large scale workflow issues.

Large executions can generate many moderation tasks.

Regular moderation review helps:

Focused workflows are easier to execute and troubleshoot.

Good examples:

Avoid overly large multi-purpose workflows when possible.

Failed tasks may occur because of:

Regularly review failed results to improve workflow quality.

Example:
A webshop imports 500 new pet food products.

Workflow:

The workflow overview allows the team to:

from one centralized location.