Enhancing Invoice Management Transparency and Efficiency Through AI Visualization

Abstract

Invoice processing systems are critical components of financial operations. They are responsible for managing the receipt, obtaining validation, or approval, and posting of vendor invoices within the Procure-to-Pay cycle. Despite having high service level agreement compliance rates, the current semi-automated invoice processing system lacks real-time visibility in individual staff productivity. This causes delays, recurring accuracy errors, and operational inefficiencies. Manual data consolidation and limitations in optical character recognition accuracy further hinder timely decision-making. Therefore, there is a need for an intelligent, automated solution to enhance transparency and streamline performance monitoring. This study adopts a Design Science Research methodology to develop and evaluate an AI-supported interactive dashboard for invoice processing within a Shared Services Centre. The operational framework consists of five structured phases: investigation, research design, data collection and analysis, dashboard development, and reporting. The proposed solution integrates AI-driven data extraction with dynamic visualisation capabilities in Power BI, guided by Shneiderman’s Eight Golden Rules to ensure usability, clarity, and human-AI collaboration. The dashboard comprises four core modules focused on processor performance, turnaround time tracker, processing time variation, and workload distribution. Findings demonstrate that the dashboard enables real-time performance monitoring, facilitates data-driven decisions, and improves transparency across financial operations. This research contributes a practical tool to address current system limitations and sets the stage for future enhancements, including integration with SAP Intelligent Robotic Process Automation (iRPA) and real-time monitoring features for proactive process management.