Artiplane · enterprise AI platform

The fashion semantic layer: one product across five systems

SAP 204417, PLM ST-1207 and POS 1207-BLK-M are the same product. Until a semantic layer says so, no answer is reliable.

Semantics

2026-09-04 · 6 min read

Key takeaways

  • Canonical entities come before any agent.
  • Colorway, season and market are business concepts, not columns.
  • One definition means one answer, everywhere.
How an enterprise RAG answer is built

ERP, PLM, CRM and documents keep provenance and version.

Identity first

A fashion product exists in every system with a different code and a different granularity: style, colorway, size, barcode.

The semantic layer declares one canonical Product and attaches each system record to it, with the mapping that produced the link.

Entities that matter

Twenty-four entities cover most fashion questions: product, colorway, season, collection, store, market, supplier, order, allocation.

Each carries the terminology the business actually uses, so a Orchestrator answers in the language of the buyer, not of the table.

Why it is not a data warehouse

A warehouse stores rows; the semantic layer stores meaning, relationships and rules, and it is read at decision time.

It is also what makes provenance possible: every figure can be traced back to the system that produced it.

FAQ

Frequently asked questions

Why does “Identity first” matter?

A fashion product exists in every system with a different code and a different granularity: style, colorway, size, barcode. The semantic layer declares one canonical Product and attaches each system record to it, with the mapping that produced the link.

How does “Entities that matter” work?

Twenty-four entities cover most fashion questions: product, colorway, season, collection, store, market, supplier, order, allocation. Each carries the terminology the business actually uses, so a Orchestrator answers in the language of the buyer, not of the table.

What should a team know about “Why it is not a data warehouse”?

A warehouse stores rows; the semantic layer stores meaning, relationships and rules, and it is read at decision time. It is also what makes provenance possible: every figure can be traced back to the system that produced it.

Continue with

  • Enterprise RAG vs fine-tuning: which one should a company choose?

    Read the article
  • Enterprise RAG evaluation: a 10-point checklist before production

    Read the article
  • AI in fashion retail: 7 processes to transform before adding a chatbot

    Read the article