AI-generated product tags for every category
Turn product images, product feed fields and API-provided product information into rich, structured metadata that powers search, filters, merchandising, SEO and recommendations.
Why metadata matters
Great product discovery starts with great product data
Customers search in detailed, natural ways. They look for style, colour, material, fit, function, occasion, sleeve length, collar type and many other attributes. If those details are missing from your product data, your products are invisible to shoppers actively looking for them.
Tag-Gen generates rich, structured metadata at scale from product images, existing feed fields, titles, descriptions, specifications and API-provided data — so every product can be found.
From product data to structured tags
Tag-Gen analyses product images and feed data together — extracting visual attributes, normalising text-based fields and mapping everything to your taxonomy through a purpose-built retail AI workflow.
-
1
Product inputs
Images, feed fields & API-provided product data
-
2
Visual & text analysis
AI extracts attributes from images and normalises text fields from your feed
-
3
Tag generation & mapping
Structured tags generated and mapped to your taxonomy and attribute configuration
-
4
Review & delivery
Tags reviewed and approved, then delivered to your platform or export format
Not just tagged. Structured, normalised and mapped to your taxonomy.
Generate attributes that make products discoverable
Tag-Gen extracts thousands of structured data points across the attributes shoppers actually search for — building a comprehensive taxonomy of visual features, functional details and category-specific terms.
Attributes generated from images. Enriched further from your product feed and API data.
Input sources
Use the product data you already have
Okkular does not rely on images alone. Retailers can provide product information through existing feeds or APIs — including titles, descriptions, supplier copy, specifications, category fields, materials, dimensions, care instructions and other available attributes.
Okkular analyses text fields alongside images to extract, normalise and enrich product metadata. This is especially useful when product images do not show every important feature, or when technical specifications are already available in the product feed.
The right tags for every product, in every category.
Every retail segment has its own taxonomy. Tag-Gen is configured per segment to extract exactly what matters — so your metadata always matches how shoppers in that category actually think and search.
Extract fit, fabric, silhouette, colour, occasion and styling cues — exactly how fashion shoppers search.
Extract construction, comfort, sole technology and style versatility — for everything from sneakers to dress shoes.
Extract material, features, functionality, dimensions and styling cues — helping shoppers visualise how it completes a look.
Extract material, finish, size, room fit, visual style and functional benefits — in the language furniture shoppers use.
Extract aesthetic, material, functionality, durability, dimensions and how a piece complements a space.
Extract application, comfort, portability, features, suitability, materials and activity relevance — for every sport and outdoor category.
Extract age suitability, developmental benefits and play value — creating safe, engaging data sets for parents.
Extract spec-accurate features, compatibility and key technical benefits — clear enough for both browsers and buyers.
Scale & consistency
Configure once. Tag every product to your standard.
Tag-Gen uses your attribute configuration to generate consistent, structured tags across every product in your catalogue. Define which attributes matter for each category, set your preferred names and values, add synonyms — and Tag-Gen applies that taxonomy to every SKU.
Configuration works at every level: segments, product ranges, categories and individual attributes. Changes apply automatically to all products in that scope — no manual rework required.
A framed botanical print featuring pink garden roses, purple iris flowers and yellow wildflowers with green foliage, displayed in a gold ornate frame against a white brick wall.
Image alt text
Descriptive alt text — generated automatically for every product image
Okkular generates natural language alt text for every product using generated metadata — describing what is visually in the image in natural language phrasing. No manual writing. No filenames. No missing descriptors/ keywords.
Human-in-the-loop
AI speed with human control
Generated tags can be reviewed, corrected and approved before they are exported or published. Your team stays in control — Okkular handles the volume, your editors handle the standards.
Generated, structured and delivered into your workflow
Rich product attributes power every surface where customers discover and evaluate products — and Tag-Gen delivers them directly into the platforms and workflows your team already uses.
Designed to fit your catalogue workflow
Tag-Gen integrates with ecommerce platforms, PIM systems and marketplace workflows. Product data can be exported or connected via API depending on your operating model.
Generated, structured and delivered into your workflow. No manual reformatting required.
Get started
Want richer product tags without manual effort?
See how Tag-Gen can enrich your product catalogue with structured, consistent metadata — generated from images, feed data and your existing product information.