Intelligent MDM Systems: Using Agentic AI and Machine Learning to Harmonize Customer and Product Master Data Across Financial Domains

Abstract

Master Data Management (MDM) aims to ensure a harmonization of business-relevant master data entities showing consistency, accuracy, and synchronization throughout different systems and processes. While companies are able to a certain extent to implement their own MDM systems leveraging capabilities of the prevailing Enterprise Resource Planning solutions, these systems tend to be not enough to master the challenges of controlling and utilizing Master Data. Additionally to the MDM capability of traditional ERP solutions, intelligent MDM systems are required that are purpose-built yet increasingly integrated within the overall technology stack. Such intelligent MDM systems harmonize customer master data and product master data, enabling a 360-degree view of business partners' transactions,, and products sold, while also providing a single source of truth to support transactional and analytical capabilities of different enterprise applications and data lakes. Harmonizing customer and product master data is the overall focus of our paper, discussing the function, motivation, challenges, and limitations of intelligent MDM systems.

As companies disentangle their technology architecture due to the need for faster innovation cycles and more agile organizational structures, and because a wider network of partnership ecosystems becomes important, the demand for MDM capabilities will further increase. However, as master data are the essence of enterprise applications as well as the usage of business- and transnational data analytics and artificial intelligence, the need for intelligent MDM capabilities goes beyond merely improving the capabilities already offered by existing ERP solutions. Therefore, we elaborate on what we determine as intelligent MDM systems and discuss how these systems contribute to customer-centric and product-centric companies. To supplement insights from an extensive literature review on MDM systems, we conduct semi-structured interviews with around 20 experts from the areas of MDM, enterprise applications, enterprise data, and business analytics.