PROPOSED PROJECT
Product Tagging & Catalog Cleanup
Organize your products and ideas by reading details from product photos and descriptions, then let your team check the suggested categories and labels before saving changes.
WHAT IS USUALLY DONE
Merchandisers manually tag each new product. Different people use different names for the same attribute, leaving filters incomplete and campaign collections difficult to build.
WHAT SOVRUN CAN DO FOR YOU
Sovrun would define an attribute vocabulary for records describing your products and ideas and process a limited batch of catalog entries. The interface shows the suggested value, the source image or text, and fields that need confirmation. Approved edits export in the existing catalog format.
How we’d deliver it
Define
Select useful attributes and distinguish visible features from facts that require documentation.
Suggest
Extract candidate values and show the evidence for each suggestion.
Approve
Review exceptions and export an auditable update rather than overwriting the catalog automatically.
Where AI helps
Vision and language models suggest observable attributes and normalize synonyms. They must not infer materials, ingredients, or performance claims without a source.
What people approve
Merchandisers review uncertain suggestions and all sensitive product attributes. Existing verified values are preserved unless explicitly changed.
What we’d need from you
- Product images and descriptions
- Current categories and rules for grouping products
- Catalog import format
What you’d receive
- Attribute dictionary
- Candidate tagging and review tool
- Approved catalog update file
How we’d test it
Compare suggested tags with a manually reviewed sample and check that the update improves real filters or collection-building tasks.
This is a proposed project, not completed client work. The examples show what we would use and produce, not measured results.
Work,