Conceptual Model for Raising Accounts Payable Accuracy Through Process Intelligence in Research Institutions

Abstract

This paper presents a conceptual model for raising Accounts Payable (AP) accuracy in research institutions by embedding process intelligence across the procure-to-pay lifecycle. The model integrates process mining, rule-based controls, and machine-learning anomaly detection with grant compliance logic to reduce mismatches, duplicate payments, and breaches. It addresses the context of universities and research hospitals, where varied funding sources, sponsor terms, and decentralized purchasing create transaction patterns and compliance risk. The model positions AP as a data-driven assurance hub connecting principal investigators, central finance, and suppliers. The architecture has four layers: first, data acquisition that unifies ERP, e-procurement, and grant management logs via standardized event schemas; second, conformance engines encoding sponsor allowability, period of performance, three-way match, and delegation rules; third, analytics and prediction that combine process discovery, first-pass-yield forecasting, vendor normalization, and exception clustering; and fourth, workflow orchestration that returns prescriptive alerts to case managers and routes exceptions to approvers for timely resolution. Methodologically, the model adopts a design-science and DMAIC hybrid. Teams baseline cycle time, touchpoints, and first-pass accuracy; mine event logs to map as-is variants; prioritize failure modes through FMEA; implement targeted controls; and measure effects with interrupted time series and segmented regression. Data quality is elevated through master-data maintenance, vendor deduplication, and invoice OCR confidence thresholds with human-in-the-loop review. Expected outcomes include higher first-pass yield, fewer late-payment penalties, improved sponsor billing, and cleaner audit trails. Leading indicators exception rate, conformance score, and rework loops feed a control chart to sustain gains, while lagging indicators write-offs, questioned costs, and audit findings confirm risk reduction. The model also incorporates equity and accessibility by simplifying small-supplier onboarding and enabling transparent status notifications to reduce inquiry volume and payment anxiety. A change-management plan aligns incentives across finance, research administration, and procurement, with skills uplift delivered through training and playbooks. This conceptualization offers a scalable blueprint aligning AP accuracy with research integrity, stewardship of public funds, and overall operational resilience, enabling institutions to realize predictable, compliant payables operations and stronger supplier relationships.