FINTECH TRANSFORMATION: AI AND RPA BOTS FOR MULTI-AGENCY PAYMENT RECONCILIATION IN ERP

Abstract

Fintech multi-agency payment reconciliation nosology in ERP systems, including SAP S/4HANA, Oracle EBS, Microsoft Dynamics 365, and PeopleSoft, has tremendous challenges based on a large number of transactions, complicated formatting of data, and manual processing. Multi-format data, including ISO 20022 CAMT.053, MT940, BAI2, EDI 820, and PDFs are common in payment reconciliation, then the auto-match rates are 50-70%, later in the reconciliation cycle, which is T + 3 to T + 10. The paper discusses the use of Artificial Intelligence (AI) and Robotic Process Automation (RPA) to enhance the performance of fintech payment reconciliation within the ERP systems of PeopleSoft and other companies. AI models are used to extract remittance data, resolve entities, and do probabilistic matching and anomaly detection using the Optical Character Recognition (OCR) and Natural Language Processing (NLP). RPA bots automate the processes of data extraction, posting to ERP, and exception handling in PeopleSoft, which consumes less manual intervention. The findings prove that the AI-RPA solution raises the auto-match rate to 88-95%, precision to 97%, and minimizes the cycle time by 60-95%. The 50-70% and 30-55% decreases in exception queues and reconciliation expenses, respectively. At α =0.05, the statistical tests support the effectiveness of the system in enhancing operational efficiency, compliance, and performance. This research gives a broad guideline on how AI and RPA can be implemented to automate and streamline payment reconciliation in PeopleSoft and other ERP platforms.