AI Enrichment Agents Explained

AI enrichment agents help you populate structured product data in SKULaunch using different types of source information.

Each agent is designed to work best with a specific kind of input, such as web pages, documents, or images. Choosing the right agent makes enrichment faster and more accurate.

AI enrichment agents support data enrichment only. They extract facts and values for attributes. They are not used for generating marketing copy or descriptions.

AI enrichment agents:
• Read source content you provide
• Look for relevant product information
• Return suggested values for specific attributes

They always work at attribute level, not product level. This means you decide which attributes you want to enrich, and the agent focuses only on those.

All results are reviewable and editable before being approved.

Product information appears in many formats. A brand website looks very different from a PDF or a product image.

Because of this, SKULaunch uses multiple agents, each optimised for a particular type of source. No single agent is best for every situation.

Using the right agent helps avoid:
• Missing values
• Incorrect assumptions
• Noisy or irrelevant results

AI enrichment agents are useful when:
• Data exists but is unstructured
• Manually extracting values would take too long
• You need to enrich multiple products consistently

They work best once products are:
• Created or imported
• Assigned to the correct family
• Using clear attribute definitions

AI enrichment is designed to assist, not replace, human review.

AI enrichment agents extract facts and values such as:
• Product type
• Materials
• Dimensions
• Specifications

Content enrichment is different. It generates written content such as descriptions or bullet points using existing attribute data.

If you are trying to create product copy, use content enrichment instead.

SKULaunch currently includes the following AI enrichment agents:

• Web Search – finds public product information online
• Page Scraper – extracts data from known product pages
• Text Extraction – structures information from text or documents
• Image Extraction – reads information from images and packaging

Each agent has its own help article explaining when to use it and what it works best with.

• AI enrichment works best with good input data
• Start with core attributes before enriching everything
• Always review suggested values
• Use AI incrementally, not all at once

AI enrichment is most effective when combined with a strong data model and clear attribute definitions.