Catalog
Product data that keeps up with your variants
Fifty products or a million, the experience is the same: schema-based data, fast faceted search and imports that just keep going.
- HX2F-010-TA10 mm 2 flute square endmill TiAlN£49.01
- HX4F-012-TA12 mm 4 flute square endmill TiAlN£58.63
Try a search
Why a catalogue
The catalogue is the ground every quote stands on
Structured product data is what turns a customer request into a matched, priced answer. Without it, the AI can only read a request; with it, the AI can resolve one.
Find anything in milliseconds
Faceted search across every typed attribute, engineered to a p95 under 300ms at 100,000 rows, so a customer question is answered while they are still on the call.
Matched, not just searched
An incoming RFQ resolves to a real product with an auditable fit rating and a price, grounded in your catalogue rather than a hopeful guess.
One source of truth
Price, stock and lead time live in one schema that feeds quotes, the configurator and servicing, so a change lands everywhere a product is quoted at once.
Fifty products or a million
The same schema, search and import behave identically from a short price list to a million-row catalogue. The experience does not change as you grow.
From PDF to structured
Most catalogues start life as a PDF
A 200-page price list or a scanned product book is a reference you read, not data you can search, match or quote from. Kabaido closes that gap.
Kabaido reads it with the same engine that reads a customer request. It pulls the PDF text layer, drops to vision on scanned pages, and serialises every table into rows, keeping the source span on each value so an extracted price or dimension can be checked against the page it came from.
Nothing is invented. A figure the page does not state stays null with a clarification, never a guess. The structured rows are yours to review and bring into a schema, where the catalogue's faceted search and matching switch on.
The same parser and the same cite-or-abstain rule run on every request, so the way a product PDF is read is the way an RFQ is read. See how the AI reads a request.
Import
From a CSV to a searchable catalogue in one import
The steps below walk the real 300-row cutting fixture that every new organisation can load as sample data. Your own price file follows the same path.
catalog-cutting-300.csv
300 rows in 20 columns
- The file streams to storage as you pick it, with the size shown as it goes.
- Delimiter, encoding and the header row are sniffed automatically.
- Spreadsheets work the same way as CSV.
Schemas
Schemas that know what a flute count is
Eight domain schema presets ship with the attributes that matter: Cutting, Workholding, Abrasives, CNC, Metrology, Coatings, Materials and Fluids. Fully custom schemas sit alongside them.
Every field is typed: number, text, select, boolean or dimension with a canonical unit. A flute count is a number, a coating is a select and a diameter is a dimension in millimetres, so 4FL, TiAlN and dia 10 all land where they belong.
Each field carries synonyms that work twice. They auto-map your CSV headers on import, so dia, d1 and ø all land on Diameter, and the same synonyms feed extraction, so the abbreviations customers write in requests resolve to the right attribute. Facet-flagged fields build the facet rail on their own, so faceted search comes straight from the schemas rather than from configuration.
Core fields such as sku, title, price and stock are implicit on every schema, and versioning is additive: removing a field archives it and the data stays. Search is engineered to a p95 under 300ms target at 100,000 rows. When a product is made to order rather than stocked, a schema is the starting point for a configurator.
Live sources
Connect the system that already holds your prices
A CSV is the easy start. When the catalogue lives in another system, connect it once and keep it current.
Kabaido reads from a REST source with simple key or token auth. You map the response to your schema once, then it refreshes on a schedule you set, so price and stock stay current without another manual upload.
For systems that push rather than wait to be read, the bulk API upserts 10,000 products per call and signed webhooks fire on quotes, orders and service events. Connections, auth and the full event list live on the integrations page.
curl "/api/v1/quotes?limit=2" -H "Authorization: Bearer kbd_..."- Q-26-00026£528.96inc VATvalid until 11 JulDraft
- Q-26-00025£202.08inc VATvalid until 10 JulDraft
Money in minor units. 600 requests per minute per key.
Operations and scale
The work around the data, from fifty products to a million
Catalogue limits run from 250 products on Free to 1,000,000 on Scale. The day-to-day operations are part of the product, not a bolt-on.
- Bulk edits, safely
- Filter to a set, change a price, a coating or a status and apply it across thousands of rows at once. The edit runs in batches with a progress bar, and every change lands in an audit trail with who, what and when, so a wrong move is easy to find and reverse.
- Lifecycle states
- Products archive rather than delete. Status is a first-class field, so a retired sku keeps its history and the quote lines that reference it.
- 3.1 certificates
- Certificate fields like EN 10204 3.1 are typed attributes on materials, treated as hard requirements when the customer states them.
- Your data stays yours
- Take a full CSV export of the catalogue at any time. No lock-in, no export fee.
Data security
Your catalogue is competitive data
Prices, costs, part numbers and customers are among the most sensitive data you hold. The catalogue is sealed the same way the rest of Kabaido is.
Every row is organisation-scoped and filtered by database-enforced row level security, so one tenant can never read another's products. Import files land in tenant-prefixed storage under the same rule.
Data is encrypted at rest, and integration credentials carry an extra application-layer encryption on top. Your catalogue is never used to train models, and the AI provider is configured for zero retention, so the prices you upload stay yours.
Hosting is in the UK. Take a full CSV export whenever you like, and ask us to delete your data at any time; on cancellation we delete or return it. Read our security posture.
Sealed by default
- Database-enforced row level security on every product row
- Tenant-prefixed storage for every import file
- Encryption at rest, plus an application layer over integration credentials
- Never used to train models, with zero retention at the AI provider
- Hosted in the UK
- Full CSV export and deletion on request, with no lock-in
Bring the price file you already have
Start free and import a CSV on any tier, then connect a live source when you are ready.