Search

Integrating AI-driven recommender systems has proven highly successful in many industries, prompting the banking sector to explore personalised client recommendations. Given the interpersonal nature of banking sales to corporate clients, wherein AI systems recommend to Relationship Managers who facilitate interactions with clients, there is a critical need for explainability in the AI-generated recommendations to support commercial activities. Our work leverages Generative AI and Large Language Models to synthesise natural language explanations for AI algorithms’ motivations, tailored for non-technical users in the banking environment. Through a case study in a major bank, Intesa Sanpaolo, our approach successfully replaces manual expert labour, offering scalable, efficient, and business-relevant explanations. Our study addresses key research questions and contributes by presenting an enriched presentation of SHAP explainer outputs in banking, validated against expert standards. We also explore the impact on the business, providing insights into the value of transparent AI-driven recommendations in the evolving landscape of banking services.

In the era of rapid growth and transformation driven by artificial intelligence across various sectors, which is catalyzing the fourth industrial revolution, this research is directed toward harnessing its potential to enhance the efficiency of decision-making processes within organizations. When constructing machine learning-based decision models, a fundamental step involves the conversion of domain knowledge into causal-effect relationships that are represented in causal graphs. This process is also notably advantageous for constructing explanation models. We present a method for generating causal graphs that integrates the strengths of Large Language Models (LLMs) with traditional causal theory algorithms. Our method seeks to bridge the gap between AI’s theoretical potential and practical applications. In contrast to recent related works that seek to exclude the involvement of domain experts, our method places them at the forefront of the process. We present a novel pipeline that streamlines and enhances domain-expert validation by providing robust causal graph proposals. These proposals are enriched with transparent reports that blend foundational causal theory reasoning with explanations from LLMs.