Automated Extraction: Turning PDF Inquiries into SAP Quotations with GenAI

In B2B commerce, prospective buyers send Requests for Quotation (RFQs) in unstructured PDF formats, spreadsheets, or long emails. Functional SD teams historically spent hours extracting part numbers, descriptions, and tender clauses manually. Generative AI orchestration on SAP Business Technology Platform (SAP BTP) has completely automated this pipeline.

Intelligent Document Processing via SAP AI Core

Using SAP Document Information Extraction coupled with Generative AI Large Language Models (LLMs), unstructured documents are ingested, OCR-scanned, and parsed. The system automatically cross-references legacy customer part numbers with the SAP Material Master (MARA) and Customer-Material Info Record (KNMT), generating a draft quotation (VA21) in seconds.

• Optical & LLM Extraction: Accurately parses non-standardized tables, hand-annotated quantities, and varied units of measure.
• KNMT Resolution: Matches customer-specific naming conventions directly to internal SAP material codes.
• Cycle Time Reduction: Cuts quote turnaround from 48 hours to under 15 minutes, boosting quotation win rates.