LLM-MAPPING

LLM-assisted EDI mapping, AI & data integration applied to EDI.

Definition

LLM mapping refers to the use of a large language model to generate an EDI mapping (X12 → JSON, EDIFACT → UBL, etc.) from a natural-language description or a sample. The LLM proposes XSLT / DataWeave / mapping engine code that the engineer verifies via Schematron unit tests. Stedi, Workato, MuleSoft offer this feature since 2023.

Origin

Approach emerged 2023 with GPT-4. Early adopters: Stedi (Map Builder AI 2023), Workato (Copilot for Mapping 2024).

Example in context

Prompt: 'Map this X12 850 line item to UBL OrderLine, with priceUnit converted from /EA to /LB if classification = 31201610'. LLM returns XSLT, engineer validates.

Related terms

• AI mapping — general concept.
• Schema inference AI — related discipline.