OPTIMIZING BANKING OPERATIONS THROUGH PROCESS MINING, BIG DATA, AND MACHINE LEARNING

Abstract

This thesis develops an integrated, governance-aware framework for improving banking operations through process mining, big-data engineering, and machine learning. Drawing on process-management literature and banking-risk standards, it explains how transactional event logs can reveal bottlenecks, how heterogeneous data can support timely decisions, and how validated predictive models can prioritize interventions. An illustrative application to payments, customer onboarding, and credit processing is proposed. The contribution is a measurable implementation architecture rather than a claim of achieved efficiency gains. The framework links operational speed, control effectiveness, customer experience, and model accountability.