Cloud-Native Financial Automation Using Agentic AI and DevOps-Integrated Generative Intelligence

Abstract

The foundations of DevOps enable rapid software development and deployment by a variety of organizations. Continuous integration, continuous delivery, infrastructure as code, and configuration as code are now mature concepts. However, DevOps techniques remain largely unexplored in the context of cloud-native financial automation. By investigating financial automation using these techniques, a framework for integrating financial automation throughout the BANK Financial Evolution through Reinvention program, especially into credit risk-related workstreams, is derived. To support such integration, continuous delivery techniques for regulated environments, leveraging the cloud-native architecture of the Program’s ecosystem, are presented. Cloud-native architectural patterns related to supporting automated financial processes are also proposed, making use of the concepts of microservices, service meshes, serverless computing, and data mesh. Furthermore, Agentic Artificial Intelligence (Agentic AI) concepts are applied to financial automation, especially in relation to agentic roles and the levels of autonomy of Agentic AI. Finally, an exploration of generative intelligence’s role in supporting and facilitating the work of human agents who undertake financial automation is undertaken by considering the automation of a variety of workflows including report generation, the creation and maintenance of compliance documentation, and providing audit trails to help support compliance and validation activities.