Enterprise workflow modernisation through artificial intelligence-driven integration of legacy banking systems

Abstract

This paper examines the role of artificial intelligence (AI) in modernising enterprise workflows within legacy-dependent banking environments. It identifies structural limitations in traditional banking systems, including siloed architectures, limited interoperability and reliance on manual processes, which constrain efficiency, scalability and responsiveness. To address these challenges, the paper proposes an AI-driven integration framework that combines machine learning, natural language processing and robotic process automation within a modular, middleware-based architecture. The framework enables intelligent data extraction, automated decision making and real-time workflow orchestration across legacy and modern systems. It introduces AI-enabled middleware connectors and agent-based orchestration to support seamless interoperability while preserving existing infrastructure investments. The study evaluates the framework across digital lending and accountopening processes, demonstrating substantial improvements in operational performance, including reduced processing times, decreased manual intervention and enhanced compliance accuracy. The findings highlight the potential of AI to transform enterprise workflows by enabling dynamic, data-driven operations and improving risk detection and regulatory adherence. The paper also addresses implementation challenges, including data governance, model interpretability and integration risks, emphasising the need for robust architectural design and explainable AI. Overall, it provides a practical, scalable approach to achieving digital transformation in complex banking ecosystems and improving longterm operational resilience. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.