ORIGINAL RESEARCH article
Front. Artif. Intell.
Sec. AI in Business
Volume 8 - 2025 | doi: 10.3389/frai.2025.1612772
This article is part of the Research TopicSoft Computing and Artificial Intelligence Techniques in Decision Making, Management and EngineeringView all 4 articles
The Role of AI-Enhanced Fast Delivery Services in Strengthening Customer Retention and Loyalty in Competitive Markets
Provisionally accepted
Apoorva Kasoju*
Tejavardhana Vishwakarma- Amazon (United States), Seattle, United States
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
This research presents an AI-enhanced framework to optimize last-mile delivery systems by integrating predictive analytics, Reinforcement Learning (RL), and customer personalization. The predictive analytics component utilized XGBoost and Random Forest models to forecast delivery times. Random Forest achieved better performance, with a Root Mean Square Error of 1.52 and an R-squared value of 0.56. RLbased route optimization improved operational efficiency by reducing the average delivery time from 31.2 to 25.4 minutes, increasing timely deliveries from 78\% to 92\%, and reducing idle time by 15\%. Customer personalization, driven by sentiment analysis and clustering, increased positive sentiment from 68\% to 80\%. It also improved Net Promoter Scores from 68 to 85 and increased customer retention from 74\% to 89\%. The proposed framework addresses the challenges of last-mile delivery by combining datadriven predictions, adaptive routing, and personalized customer strategies. Future work will explore realworld implementation using real-time traffic data and advanced personalization techniques to improve adaptability and scalability.
Keywords: AI-enhanced delivery, predictive analytics, reinforcement learning, customer personalization, Last-mile delivery, Operational efficiency
Received: 16 Apr 2025; Accepted: 29 Aug 2025.
Copyright: © 2025 Kasoju and Vishwakarma. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Apoorva Kasoju, Amazon (United States), Seattle, United States
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.