Case study 01 / WB / OZON · NDA

From product data to a finished listing image.

A WB and Ozon seller workspace combines catalog work with content preparation. One editor brings the product image, copy and visual layout together.

Audience: Marketplace sellers and teams responsible for catalog maintenance and listing content.
Scope: Seller workspace, listing and review monitoring, visual generation, manual editing and export.
Stack: WB / Ozon · AI generation · Product card editor · Dashboard

Overview

The problem behind the interface.

The brief

Bring a seller's daily work into one place: product range, orders, listing monitoring and image creation.

Scope of the work

A listing image needs more than a suitable photograph: its headline, specifications and output size must also work together. These steps are arranged around one preview so a seller can refine a chosen variation into a finished asset.

Customer journey

A clear route through the product.

Choose a product

Catalog work and listing tasks share the seller workspace.

Build a visual

Choose a visual treatment and generate image variations.

Refine the content

Edit the headline, specifications and text placement beside the preview.

Export the asset

Download the composition in the selected format for subsequent publishing.

Design decisions

The decisions that shape the product.

Keep the composition in view

The source product sits on the left, the listing image in the centre, and copy controls on the right. Each edit has a clear relationship to the image. A seller can check the headline, selected specifications and text legibility while keeping the full composition visible.

Make generated content editable

A generated variation is a starting point for refinement. The editor supports image replacement and changes to copy, colour, size and placement. Updating a specification or headline does not require recreating the entire visual, and the seller retains control over the content of the listing.

Connect content to daily catalog work

The editor belongs to the same workspace as the assortment, orders, monitoring and reviews. Listing content becomes part of the store's everyday workflow. WB and Ozon selection, image dimensions and export controls are available directly in the editing screen.

Screens

A closer look at the interface.

Source, composition and controls

The product and visual treatments sit on the left, the preview in the centre, and copy on the right. This English demonstration preview uses generated product imagery and fictional specifications.

Refinement before export

Specification fields, typography controls and the export format share the right panel. This is a closer reading of the same demonstration screen: review the content, then prepare the asset for download.

A complete listing composition

A photograph, headline and key product properties form one visual asset. This is a demonstration output with generated imagery and fictional copy and numbers, rather than a published client listing.

Outcome

What we built.

Implemented a seller workspace and listing content editor, covering visual selection, manual refinement and export. The portfolio uses demonstration product imagery and fictional specifications; it does not show seller account data or sales results.

Discuss a similar project

Where this approach fits

Useful for marketplace listings, product catalogs and branded image tools that need visual variations and precise manual editing.

WB / OzonAI generationProduct card editorDashboard

Project under NDA. The image contains no sales data, accounts or seller identifiers.

Next case studyAI agent with human oversight