Year: 2024
Type: article
Abstract: Traditional Business Intelligence and Analytics in customer relationship management (CRM) and the service industry are essential for delivering exceptional customer service and ensuring prompt resolution of warranty claims. As customer service and warranty management processes evolve, maintaining the integrity of warranty claims becomes increasingly critical for preserving business trust and financial stability. This paper investigates the application of advance analytics and machine learning techniques to develop anomaly detection models that identify fraudulent patterns in warranty claims data. By analyzing customer service interactions and contract details within a CRM system, this study employs AI, data mining techniques, the Isolation Forest algorithm, and K-Means clustering to detect anomalies based on temporal and geographical patterns in claims. The methodology focuses on key performance indicators (KPIs) such as the timing of Annual Maintenance Contract (AMC) purchases and the clustering of claims by geographic locations associated with sales representatives and other business processes. The findings reveal significant correlations between fraudulent claims, submission timing, customer history, interactions and specific geospatial patterns, indicating potential collusion between sales representatives and customers. This paper contributes to the detection of malpractice in service-oriented industries and provides valuable insights for businesses seeking to enhance their warranty management processes through advanced data analytics.
Author: Savio Dmello
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4402990474