AI Product data, structured. Air tools · HVAC · plumbing · electrical · workshop equipment Trade buyers search by spec. Make sure the spec is there.
"3/4 drive 1,000 Nm", "1/2 inch BSP inlet", "415 V three phase" — the queries that convert are specifications. Mavitra AI pulls capacities, ratings, connections and standards out of your product pages and spec sheets and writes them where filters and feeds can use them.
Pit-lane impact wrench, 1" square drive, magnesium housing with carbon fibre front cover. Free speed 9,000 rpm. Torque range 500–2,900 Nm. Supplied with right-hand identification ring.
Drive size, torque and speed became filters on the impact-wrench collection; the same values fill the merchant's own blank spec fields, so nothing is shown twice.
Your spec sheets are the evidence. We just read all of them.
One schema per category
Air supply, hand tools, hoists, pumps, ventilation — each category has its own attribute set with units, so an FRL kit is never asked for torque.
Your specs win
Values already on your product (your own metafields, your ERP feed) are treated as the strongest evidence. We never contradict a stated spec.
Units, fractions, ranges
1/2" and 13 mm are the same hose. "500–2,900 Nm" is a min and a max. The comparer understands the formats trade catalogs actually use.
Fill your blanks, never duplicate
Where your own spec field is empty and we have the fact, we fill it. Where you already have it, we leave it. A buyer never sees the same fact twice.
139 attribute definitions across 20 categories
Spec filters
Drive size, torque, inlet, pressure — as Search & Discovery filters on the collections that need them.
Shopping feed
Structured attributes and product types so a search for "1 inch impact wrench 2000nm" matches your SKU.
Backfilled spec fields
Your own definitions filled where they were blank — exportable to your ERP or print catalog.
Questions from trade suppliers
We already have specs in our ERP feed. What do you add?
Coverage and consistency. On our first live catalog, 1,423 of 1,599 products had at least one spec field blank that the product page itself stated. We filled those; we did not touch the ones you had.
How do you measure accuracy on our catalog?
Against your own data. Every extracted value with a matching field in your specs is compared automatically — unit-aware — and reported per attribute before anything is written.
Our Shopify store is near the metafield definition limit.
So was our first customer's (238 of 256). Only filterable attributes get definitions; everything else is written as values. We check headroom before installing anything.
Which brands and categories have you done?
Air tools, hand tools, hoists, pumps, ventilation and workshop equipment across 60+ brands on one live catalog. Each new category gets its own schema before we run it.
Find out which specs your buyers can't filter on
The free audit enriches 100 of your products and lists every spec your pages state but your store can't search.