APPLICATION OF GENERATIVE AI IN ENTERPRISE DATA MIGRATION: A CASE STUDY ON SAP SYSTEM TRANSFORMATION

One of the most challenging and resource consuming parts of the implementation of large scale SAP system transformations, especially during switches from SAP ECC to SAP S/4HANA, is enterprise data migration. This paper shares an applied case study with regard to the use generative artificial intelligence (GenAI) and/or large language models (LLMs) for automating data profiling, schema mapping, and the generation of transformation rules as well as reconciliation reporting, within the context of the data migration life cycle of a simulated enterprise transformation program. A process benchmarking/quantitative defect analysis approach was used over five simulated mock-migration cycles. The findings show that the GenAI-assisted approach resulted in a reduction of the total migration effort by around 53%, a decline of Average Defects per Mock Cycle by 65.5%, and an increase in 1st pass data mapping accuracy by 21% from 71% to 92% to cut over readiness by almost 41%. The results indicate that GenAI has great potential to significantly complement, not replace, human data migration experts by handling repetitive tasks and pattern recognition but maintaining human involvement in governance and critical decisions. The implications for enterprise architects, SAP practitioners, and the future research on responsible AI adoption in enterprise resource planning (ERP) transformation are discussed.