“Explainable AI (XAI) in Salesforce for Transparent Decision-Making Systems"
Power System Technology · 31.12.2025 · 49 (4) · 4372-4409
DOI: 10.52783/pst.3453 ↗Kurz gefasst
Fachartikel in der Zeitschrift Power System Technology, online erschienen am 31.12.2025. Verfasst von Varun Anish Reddy Yeruva.
Abstract (Originalsprache)
Background:The rapid adoption of Artificial Intelligence (AI) within customer relationship management (CRM) ecosystems has significantly transformed enterprise decision-making processes. Among enterprise platforms, Salesforce has emerged as a dominant AI-enabled ecosystem through tools such as Einstein Analytics, predictive lead scoring, recommendation engines, customer segmentation, and automated workflow optimization. However, the increasing complexity of machine learning algorithms has introduced challenges associated with algorithmic opacity, decision bias, regulatory compliance, and user trust. Explainable Artificial Intelligence (XAI) has therefore become essential for enhancing transparency, interpretability, accountability, and ethical governance in enterprise AI systems. This study investigates the integration of XAI frameworks within Salesforce-driven intelligent decision-making systems and evaluates their impact on predictive performance, organizational trust, interpretability, and operational efficiency. Methods:The research employed a hybrid quantitative–experimental methodology using a simulated enterprise CRM dataset comprising 52,486 customer interaction records collected from sales, marketing, and customer support domains. The study implemented three machine learning models within a Salesforce-compatible analytical architecture: Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Deep Neural Networks (DNN). To enhance interpretability, XAI techniques including SHAP (SHapley Additive exPlanations), LIME (Local Interpretable Model-Agnostic Explanations), and feature importance mapping were integrated into the predictive framework. The models were evaluated across key performance indicators including prediction accuracy, precision, recall, F1-score, Area Under Curve (AUC), interpretability index, user trust score, and decision transparency metrics. Statistical validation was performed using ANOVA, Pearson correlation analysis, and paired t-tests with significance measured at p < 0.05. Results:The experimental findings demonstrated that the XGBoost-XAI integrated model achieved the highest predictive performance with an accuracy of 93.8%, precision of 91.6%, recall of 92.4%, F1-score of 92.0%, and AUC value of 0.957. Incorporation of SHAP-based explainability improved model interpretability by 41.3% compared with conventional black-box AI systems. User trust and managerial acceptance increased significantly from a baseline mean score of 3.1 to 4.5 on a 5-point Likert scale (p < 0.001). Feature attribution analysis identified customer engagement frequency (24.7%), response time (18.9%), historical purchase behavior (16.4%), and sentiment polarity (12.8%) as the most influential decision-making variables. Furthermore, explainability-enabled models reduced perceived algorithmic bias by 28.6% and improved regulatory compliance readiness by 35.2%. Correlation analysis revealed a strong positive relationship between interpretability and user trust (r = 0.84), indicating that transparent AI explanations substantially influence organizational confidence in automated decisions. Conclusion:This study establishes that Explainable AI significantly strengthens the effectiveness, transparency, and reliability of Salesforce-enabled intelligent decision-making systems. The integration of interpretable machine learning models not only enhances predictive accuracy but also improves user trust, fairness perception, and organizational acceptance of AI-driven decisions. The proposed framework provides a scalable and practically implementable model for enterprises seeking ethical and transparent AI adoption within CRM ecosystems. The research contributes to the growing body of knowledge on enterprise XAI by presenting statistically validated evidence supporting the role of explainability in sustainable and accountable AI governance for next-generation digital enterprises. DOI : https://doi.org/10.52783/pst.3453
Zitieren
APA
Yeruva, V. A. R. (2025). “Explainable AI (XAI) in Salesforce for Transparent Decision-Making Systems". Power System Technology, 49(4), 4372-4409. https://doi.org/10.52783/pst.3453
Vancouver
Yeruva VAR. “Explainable AI (XAI) in Salesforce for Transparent Decision-Making Systems". Power System Technology. 2025;49(4):4372-4409. doi:10.52783/pst.3453
BibTeX
@article{Yeruva2025,
author = {Yeruva, Varun Anish Reddy},
title = {“Explainable AI (XAI) in Salesforce for Transparent Decision-Making Systems"},
journal = {Power System Technology},
year = {2025},
volume = {49},
number = {4},
pages = {4372--4409},
doi = {10.52783/pst.3453}
}
EUVIMED