Survey of Agents Methodologies in the Financial Domain

Download PDF
Open Webpage

Mohammad Luqman, Himaanshu Gauba, Prabhat Kumar, Akshar Prabhu Desai, Ajay Yadav, Ritu Prajapati, Pranjul Yadav

Published: 2025, Last Modified: 24 Mar 2026IEEE Big Data 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Agentic AI techniques are autonomous systems that can make their decisions independently to work towards a specific goal in a dynamic environment. This paper explores the emerging field of Agentic AI in finance, examining the opportunities it presents through automation, faster decision making, and personalization along with the challenges. In addition, we also discuss the various techniques used to develop such agents along with their inherent design patterns, composite structures, and frameworks used to develop and evaluate such agents. Lastly, we discuss the future and emerging directions within this research area by focusing on multiagent and human-agent collaboration. We believe that this manuscript will provide a comprehensive overview of work done in the cross-disciplinary area of Agentic-AI and finance.