Explainable AI (XAI) Frameworks for Automated Sentiment-Based Ticket Prioritization in Enterprise Customer Relationship Management (CRM)

• Abey Litty

  Texas University

  Author

Enterprise Customer Relationship Management (CRM) systems increasingly face the challenge of managing overwhelming ticket volumes while maintaining service quality and customer satisfaction. Traditional rule-based prioritization methods fail to capture nuanced customer sentiment, leading to delayed responses for urgent issues and diminished customer trust. This study addresses the critical gap between accurate sentiment prediction and transparent decision-making by proposing an Explainable AI (XAI) framework for automated ticket prioritization. The research employs a design-based methodology integrating Natural Language Processing (NLP) with ensemble machine learning models, specifically BERT for sentiment classification and XGBoost for priority scoring, augmented with SHAP (SHapley Additive exPlanations) for model interpretability. The proposed framework achieved an overall classification accuracy of 89.4% in sentiment detection, with a 23.7% reduction in average first-response time compared to baseline methods. The XAI integration enabled stakeholders to understand prioritization rationales, improving trust and adoption rates. This research contributes a replicable framework that bridges predictive performance and operational transparency, offering practical implications for CRM administrators seeking to optimize support workflows while maintaining accountability in automated decision-making.

• pdf

Copyright (c) 2026 Abey Litty (Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.