Volume 16, Number 1

The Future of Financial Assistance: Leveraging LLMs for Personalized, Human-Like Interactions

  Authors

Hamza Landolsi ¹ and Ines Abdeljaoued-Tej ², ¹ University of Carthage, Tunisia, ² University of Tunis El Manar, Tunisia

  Abstract

Generative Artificial Intelligence (GenAI) is transforming the business landscape by enhancing accessibility, efficiency, cost-effectiveness, and innovation. This paper investigates the application of Large Language Models (LLMs) and GenAI in the financial sector, proposing a novel framework to reimagine robo-advisory systems. The framework shifts from traditional, rigid platforms to a more humanized approach that actively engages investors in a personalized asset selection process while leveraging LLMs to better understand their goals and profiles. We present an end-to-end solution designed to address key limitations of conventional roboadvisors, such as inflexibility, restricted asset type offerings (typically limited to equities), and challenges in accessing high-quality, real-time data. The proposed architecture incorporates dynamic client profiling, risk aversion estimation, and portfolio optimization. A tailored asset selector agent, supported by robust data pipelines, ensures the curation of up-to-date market information. Through iterative development, we utilized prompt engineering and multi-agent workflows to refine user interactions and deliver actionable insights. By implementing an innovative chatbot platform, we demonstrate the potential of LLMs to revolutionize customer service, enhance investor engagement, and provide strategic financial guidance. This study highlights the transformative impact of GenAI in creating more adaptive, personalized, and effective financial advisory solutions.

  Keywords

Generative AI, Large Language Models (LLM), Big Data, Practical Applications, Agentic Design Patterns, Finance, Investment analysis, Portfolio Optimization.