Use case · Product catalogues

One T-Box, many taxonomies.

The reference example: a Cars catalogue with five schemes — body styles, fuel types, market segments, manufacturing geography — all sharing the same classes. Built for retail, automotive, telecom, life sciences.

The problem

Your catalogue lives in three tools and none of them know it.

Product information sits in your PIM. Categories sit in your CMS. Marketing facets sit in a Google Sheet. Each tool models part of the catalogue; none has the typed graph that connects them. Adding a new body style means three change requests across three vendors and two weeks of cross-checking.

The fix isn't a fourth tool. It's a single typed graph that hosts every taxonomy your team needs, all sharing the same classes — so when a Manufacturer is added, every scheme that uses Manufacturers can reference it without a migration.

The shape

One T-Box, every scheme.

The Cars-style ontology defines nine classes once: Manufacturer, Model, Engine, BodyStyle, FuelType, DriveType, Segment, Country, Transmission. Five schemes layer on top, each curated independently — model catalogue, body-style taxonomy, fuel-type taxonomy, market segments, manufacturing geography.

  • Each scheme has its own broader/narrower hierarchy
  • Cross-scheme relations stay intact (a Model belongs to a Manufacturer in any scheme)
  • Switching schemes in the UI is one click
  • Search and bulk actions stay scoped to the active scheme

In the editor

Curate every taxonomy from one place.

The Taxonomies tree shows every scheme in the same workspace. Drag-drop reparenting works inside a scheme. Cross-scheme moves are deliberately blocked — they'd imply a semantic redefinition, not a hierarchy tweak. The change history is shared across schemes so a single tag captures every edit.

  • Scheme switcher in the left rail · keyboard accessible
  • Per-scheme breadcrumbs and counts
  • Tag once, capture every scheme's state
  • Export per scheme (SKOS Turtle) or full ontology (OWL XML)

Industries that fit

The shape repeats everywhere.

Retail

Products by department, by season, by sustainability tier, by price segment. Same SKUs; four scheme views.

Automotive

Models by manufacturer, by body style, by fuel type, by market segment. Five schemes, one T-Box (the reference example we ship).

Telecom

Plans by tier, by family, by region, by use case. The cross-scheme relations are the value — a Plan's Family applies in every market view.

Pharma

Compounds by therapeutic area, by mechanism, by trial phase, by molecule class. Curators specialise; the T-Box is shared.

Media

Content by genre, by audience, by mood, by editorial framing. The same library, four entry points for downstream apps.

Insurance

Risks by line of business, by jurisdiction, by perils, by reinsurance tier. Underwriters work in their scheme; portfolio managers see the union.

Your first week

Catalogue → ontology → API.

  1. 1

    Day 1 · Import your master catalogue

    CSV wizard maps your existing columns to ConceptClass attributes. The dry-run preview catches missing required fields before anything writes.

  2. 2

    Day 2 · Build your first scheme

    Create a hierarchy from your data. Drag-drop reparenting on the tree view. Tag v0.1 once it stabilises.

  3. 3

    Day 3 · Add a second scheme

    Same T-Box. New scheme. The Concepts already exist — you're just curating a different view of them.

  4. 4

    Day 4–5 · Wire downstream

    Ship the API call to your e-commerce, search, BI, or LLM pipeline. SKOS Turtle for search, JSON-LD for RAG, CSV for analysis.

  5. 5

    Day 6–7 · Invite the curators

    Per-scheme curators get edit rights. The validation panel keeps everyone honest. Tags continue to roll forward.

Catalogue. Curate. Connect.

The Cars starter is loaded by default — clone it, rename it, replace the data.