Skip to content
Machine-ready product intelligence

One catalog. Every channel reads it differently. Now they all get what they need.

Google, Meta, TikTok, Pinterest and AI assistants each read products in their own format — what we call a protocol. We enrich your catalog once and deliver it in the shape each one expects, built from your own verified product data, never invented.

Powered by the Semantic Commerce Engine — part of the Semantic Commerce Layer™

What every protocol receives

One product, delivered — machine-readable, in every protocol's shape.

A real product from a catalog — brand withheld. The enriched record each destination receives, built from the store's own verified evidence.

{
"brand": "a leather goods brand",
"product": "a handmade leather tote",
"category": "bags_luggage › handbags_general",
"product_type": "tote bag",
"price": 246,
"availability": true,
"use_cases": ["evening_out", "everyday", "gift"], // contexts the shopper asked for
"purchase_signals": ["genuine_leather", "handmade"], // filters that decide the buy
"functional_intent": ["hands_free_carry", "lightweight_carry", "organized_storage", "characterful_design"], // what the product is for
"functional_attributes": { "design": "vintage", "texture": "rich" },
"materials": ["leather", "bullhide"],
"search_synonyms": "+5 curated", // magnitude only — the curated terms are IP
"grounded": true // every field traces to the merchant's real catalog
}

Your catalog was written for shoppers. Machines read something else.

A product page is built for a person scrolling: photography, tone, layout. The information a machine needs — category, material, capacity, compatibility, fit, use case — is often missing, buried in a tab, or written as prose instead of structured fields.

Traditional feed tools move what you already have. If the attribute isn't in your catalog, it isn't in the feed either. So the product shows up thin in Google Shopping, gets skipped in Meta's dynamic ads, and doesn't come up at all when someone asks an AI assistant for exactly that item.

The fix isn't another mapping tool. It's producing the product information that was never written down — and then delivering it everywhere.

Built for stores selling across more than one surface.

  • Merchants running Google Shopping, Meta or TikTok catalogs that underperform
  • Stores whose product attributes are thin, inconsistent or missing
  • Merchants preparing for AI assistants and agentic shopping
  • Agencies managing catalogs across several channels and clients

Works with Shopify, WooCommerce, Wix and Squarespace. Custom architectures: talk to us. Not a fit if your catalog is already complete and you only need to reformat it — that's a mapping tool, and that's a different product.

What we need from you.

  1. 1.Your store URL. That's the starting point — we read your published catalog.
  2. 2.Read access, only if your platform requires it — many stores need nothing beyond the URL.
  3. 3.Which protocols you want to speak: Google, Meta, TikTok, Pinterest, AI assistants.

No spreadsheets, no field-by-field mapping, no theme changes.

How we handle your data

Your enriched catalog lives on NSOLVIA infrastructure and is served only to you and the destinations you choose.

  • Each merchant gets a private, rotatable access key — revocable at any time
  • Keys are stored hashed; the full key is shown to you once and never logged
  • Strict tenant isolation: your key returns your catalog and nothing else
  • Read-only relationship with your store: we don't write to your catalog in this product
  • Stop your subscription and the delivery stops — your store is untouched

What the enrichment does.

Reads your published catalog — including product content that lives in tabs, accordions or apps, where most tools stop looking.

Resolves what each product actually is: category, product type, and the attributes that matter for that category.

Extracts the evidence a machine needs to match a product to a shopper's intent — material, capacity, compatibility, fit, use case — from what you already published.

Delivers the result in the shape each destination expects: XML where XML is expected, JSON where JSON is expected, and the AI-readable formats assistants consume.

Keeps it current. Prices, stock and new products refresh on their own — the delivery stays live without you rebuilding anything.

Rule we don't break: Verified, never invented. If a fact has no evidence in your catalog, the field stays empty. Warnings and compliance text are preserved verbatim.

What you get.

  • One private URL per protocol — paste it into the destination and you're live
  • Products enriched with the attributes each destination actually reads
  • Machine-readable descriptions built from your verified evidence
  • Automatic refresh: price, stock and new products stay current
  • A dashboard where every protocol you own is listed, previewable and copyable

Add a protocol whenever you're ready. The catalog is enriched once — each new destination is a new way of speaking, not a new project.

Three steps. Minutes, not weeks.

Step 1 — Paste your store URL

We read your catalog and identify what's there.

Step 2 — We enrich it

Your products are resolved and enriched from your own published evidence.

Step 3 — Copy your URL and go live

One URL per protocol. Paste it where the destination expects it. It stays current on its own.

Want to see it before deciding? Start with 10 products, free — your products, live, in a real protocol.

The same product, told two ways.

One product from a real catalog — brand withheld — the way it reads today, and the way a destination receives it.

Before — the human catalog entry
title, one image, and a description written for a person — no resolved category, no structured attributes, no machine-readable fields.
After — what a destination receives
resolved category · structured attributes from verified evidence · machine-readable description · the format each protocol expects. (See the full record above.)

What we claim: your catalog arrives complete and interpretable at every destination you connect, and you can see exactly what each one receives before going live.

What we don't claim:we don't promise rankings, placements or sales. Those are different outcomes and they need different measurement.

Questions & answers

What merchants ask.

How is this different from the product feed I already have?
A regular feed moves the data you already have; it can't add what was never written down. If your catalog is missing category, material, capacity or use case, a mapping tool sends an incomplete product to the destination. AI-Ready Commerce Feeds™ produces that missing product information first — from evidence already published in your own catalog — and then delivers it in the format each destination expects.
What is a “protocol”?
A protocol is one destination's way of reading products — Google, Meta, TikTok, Pinterest, or an AI assistant. Each one expects a different structure, and each one is a separate delivery from the same enriched catalog. You choose which protocols you want to speak; adding another doesn't mean re-enriching your catalog.
Where do the attributes come from? Can the AI invent things about my products?
They come from your own published product information — titles, descriptions, specs, variants, materials, warnings. If a fact isn't supported by evidence in your catalog, the field stays empty rather than being filled with a guess, and warnings and compliance text are preserved word for word. Verified, never invented.
Do I need to rebuild my catalog or write code?
No. You paste your store URL, we read your published catalog, and you get one URL per protocol to paste into that destination. No spreadsheets, no field-by-field mapping, no theme edits, and no developer time.
How long until my feed is live?
Minutes for the free plan, and the same day for a full catalog in most cases. You paste your store URL, we read and enrich what's published, and you get a URL to paste into the destination — there's no mapping project, no onboarding call and no developer queue in between.
Does it stay current when I change prices or add products?
Yes — that's what the subscription is. Prices, stock and new products refresh on their own, so the URL you pasted keeps serving current data without you rebuilding anything. Maintenance isn't an add-on here; it's the product.
Can I try it before paying?
Yes. The free plan enriches 10 products of your catalog and delivers them in one real protocol — your products, live, not a demo store. When you're ready, upgrade and the rest of your catalog follows.
Who can access my catalog data?
Only you and the destinations you connect. Each merchant gets a private access key that you can rotate or revoke at any time; keys are stored hashed and never logged, and isolation is enforced so a key only ever returns its own catalog. This product reads your store — it doesn't write to it.
What happens if I cancel?
Delivery stops and your store is untouched. The enriched catalog lives on our infrastructure, so when the subscription ends the URLs stop serving — nothing is removed from or changed in your own store, because we never wrote to it in the first place.
Which ecommerce platforms do you work with?
Shopify, WooCommerce, Wix and Squarespace are supported directly. What we read is your published catalog, so what matters is that your products are live on the web — if you're on a custom build, talk to us and we'll tell you straight whether we can read it.

See it with your own products.

Start with 10 products, free — one protocol, live, no card required. Or run the free Agentic Catalog Readiness Audit™ first to see where your catalog stands today.

  • Part ofSemantic Commerce Layer™NSOLVIA's framework for practising IEO on commerce data — the interpretability layer between catalogs and the systems that read them.
  • Goes deeper
  • Alongside it

AI-Ready Commerce Feeds™sits inside NSOLVIA's interpretability stack — an open discipline, the framework that implements it, and the instruments that apply it.