Buy Vue.ai if your attributes live in the photo and you want AI on-model imagery; note its catalog stack sits inside a fintech acquirer since 2025, reads images rather than supplier spec documents, and cites no sources.
Both claim to enrich product data. This page is about where that claim stops.
Product data is a practice, not a project — and against an enrichment-AI platform the argument is about what happens after the first pass: the SKUs the run flags for review, the tags that drift the week merchandising renames a category, and the fields whose source document nobody routed into the run.
Mine every spec from every source.
Every value traced to a document you can open. The catalog is only as honest as what it was built from.
Aim the catalog at the buyer who actually buys.
Grounded data still loses if it answers questions nobody asked. Alignment is what turns specs into conversion.
Product data is a practice, not a project.
Markets move, suppliers reissue, buyers change what they ask for. A catalog that is right in March is wrong by August unless something is watching.
Scored against public documentation. Grouped by the three acts — so you can see which ones Vue.ai (Mad Street Den) leaves on your desk.
Images, labels, OCR text, EAN lookups; not supplier spec docs
PDFs, spec tables, drawings, manuals, images, sites
Custom taxonomy builder; customer defines the attribute set
Proposes fields your schema never had
Tags against preset taxonomies; no versioned vocabulary governance shown
Normalizes and governs allowed values, versioned
Auto category prediction into custom taxonomies; core strength
Auto-classifies; channel and marketplace mapping
QA dashboard only; no source citation per value
Every value cites its source doc and page
Descriptions in any tone; attributes not persona-differentiated
B2B specifier and B2C shopper enriched differently
Assortment analysis against user interest; reports to merchandisers
Reviews, search, competitor rails, social — fed back
GenAI titles and descriptions in configurable tone, style
Original copy per persona and channel
VueModel generates on-model photos from mannequin shots; fashion-focused
Generates studio-grade imagery for photoless SKUs
Batch tagging runs; engine learns from QA inputs
Re-enriches on its own after go-live
QA dashboard tracks tag accuracy; no completeness scoring
Scored against your standards; nothing publishes below bar
Exports to PIM, DAM, CMS via API or CSV
Writes back to PIM, ERP, warehouse, commerce
REST API and SDK; no MCP server found
API, webhooks, and MCP servers
Software plus QA dashboard; customer team owns output
Anglera owns the work; review is a guardrail
“Buyer signals” is the emptiest phrase in this category, so here is the literal thing. Each of these is an observation from a live market, the gap it exposes, and the field that gets created as a result.
"2e wide", "4e running shoe", "wide fit d width" — recurring queries landing on zero-result or full-category pages
Width is a fit fact, not a visual one: a photo shows the shoe, not whether the last is D or 2E. It lives in the supplier's size grid, so it reaches the tag set only if someone routes that grid in and maps its labels to codes — and reaches the facet only if the taxonomy has a field waiting for it.
shoe_width_codeUS width codes: AA, A, B, D, 2E, 4E, 6E — one value per variant, mapped from the supplier's own grid ("Wide" → 2E, "X-Wide" → 4E)"It's tagged for a 15\" laptop but my 15.6\" ThinkPad won't go in — the padded sleeve is about 14.5\" across at the opening."
An image supports "laptop compartment: yes". The buyer is deciding on interior sleeve dimensions and the largest device that actually fits — two governed numeric fields whose source is the tech pack. Someone has to decide those fields exist, fix the units, and get the tech pack into the pipeline.
laptop_sleeve_interior_dimensions_in / max_device_diagonal_inInches to one decimal, always W x H x D measured at the sleeve opening; max_device_diagonal_in as a single number, never a marketing size classA block of dress variants disapproved: "missing size_type", "invalid size_system" — the offers carry "Petite 8" in a free-text size field and nothing else
The tagging pass produced good on-page facets — midi, floral, sleeveless. The channel wants two governed enum fields with Google's own allowed values. Pre-set retail tag sets and custom tags both exist; what has to be built and kept is the parse from a legacy free-text string into those enums, and an owner who maintains it as the spec changes.
size_system + size_typesize_system: US, UK, EU, DE, FR, JP, IT, BR, CN, AU. size_type: regular, petite, plus, tall, big, maternity — parsed out of the legacy free-text size string, not appended to itA tagging run leaves a catalog in three states. Fields it filled from what it could read. Fields where the answer came back borderline and someone has to look. And fields whose source is a document nobody routed in — the cordless drill's battery platform, the downlight's field-selectable CCT, the backpack's interior sleeve width, each sitting in a spec sheet or tech pack rather than the shot. Vue.ai's Data Hub auto-retrains models on user feedback, which means somebody has to keep giving it feedback: a queue, an owner, an hour a week. That queue is what quietly stops. Then merchandising renames a category, the taxonomy shifts under the tags, and last quarter's output is confidently wrong against the new tree. The model did its job. The practice around it lapsed.
Values are normalized into a governed, versioned set of allowed values — so a filter works, and keeps working after the next import.
3/4 in0.75"3/4"19mm3/4 inchDN200.75 in (DN20)Six suppliers, six spellings, one physical size. Filters only work once they agree.
BlkblackBLACK MATTEMatte BlkRAL 9005Black — MatteFree text makes a colour filter useless. A governed value makes it a facet.
SS316316 StainlessStainless Steel 316A4 StainlessStainless Steel — 316 / A4Same alloy, four vocabularies, plus a trade name. Buyers search all of them.
Filling the fields you defined has an invisible ceiling: a catalog can hit 100% complete and still miss the attribute that loses the sale, because completeness is measured against a schema someone drew years ago. Schema Foundry reads competitor listings, buyer searches, review complaints and your supplier docs, and proposes the fields you never defined — which is where Vue.ai (Mad Street Den) stops.
How Schema Foundry works
Vue.ai, from Chennai-based Mad Street Den, sells an enterprise AI platform whose retail modules include VueTag (automated product tagging via image recognition, NLP and OCR against custom taxonomies) and VueModel (AI-generated on-model fashion imagery from mannequin, ghost-mannequin or 3D input). VueTag extracts visual attributes, predicts categories, generates titles and descriptions, and exports to PIM, DAM or CMS via API or CSV. Mad Street Den was acquired by M2P Fintech in March 2025 in a reported $10-15M distress deal, and vue.ai now presents as a horizontal AI orchestration platform spanning retail, financial services, insurance, healthcare and logistics.
Pricing: Not publicly disclosed. No pricing page, tiers or per-SKU rates are published; retail product pages route to "Request Demo". Historic material referenced a self-serve VueTag app with API access gated to "VueTag Pro", but the vuetag.ai domain and its docs site no longer resolve, so that tier cannot be confirmed as current. Enterprise deals appear custom-quoted by scope and volume.
Vue.ai (Mad Street Den) websiteFashion and apparel retailers needing visual attribute tagging at scale plus AI on-model imagery from mannequin shots — categories where attributes live in the photograph, not in a spec sheet.
We'd rather tell you here than in month three of an implementation.
Capability verdicts reviewed against Vue.ai (Mad Street Den)'s public documentation on July 14, 2026. Vendors ship quickly — if something here is out of date, tell us and we'll correct it.
Bring one category and your supplier files. In 30 minutes you'll see it enriched — complete, structured, and consistent enough to launch on — plus the attributes your schema didn't have yet.
Book a demo