The Artificial Intelligent Loan Marketing System: Integrating Predictive Modeling and Generative AI

The implementation of International Financial Reporting Standard 17(IFRS 17)  in 2023 by the In-ternational Accounting Standards Board(IASB), prompting the insurance industry to explore more efficient ways of managing asset and liability management(ALM) and predicting policy loan de-mands. This study presents the Artificial  Intelligent Loan Marketing System (AILMS), a framework consisting of three key components that integrate predictive modeling and AI services to enhance loan marketing within the insurance industry. The system first utilizes a Gradient Boosting Decision Tree (GBDT) model, analyzing 355 variables spanning customer, policy, and financial data to pre-dict policy loan demand over the next three months. This model achieved an AUC of 0.9 and a lift ratio of 7.8, marking a substantial improvement in identifying loan-prone customers. Additionally, a Text2SQL function enables marketing planners to query model results across customer segments using natural language without programming skills. The final component integrates generative AI, automatically creating personalized marketing content tailored to distinct customer segments. By intelligently streamlining key steps in the marketing workflow, the AILMS significantly enhances process efficiency, allowing insurers to rapidly generate and deploy targeted content. This AILMS provides insurers with a no-code, data-driven solution that aligns with IFRS 17, enhancing ALM and supporting precision marketing strategies. By combining predictive accuracy, accessible data querying, and automated, audience-specific content creation, the system empowers insurers to effi-ciently target and engage high-potential customers, achieving both compliance and marketing opti-mization.