Applying Large Language Models to Build the Tourmate Smart Travel Support Platform

The integration of Large Language Models (LLMs) into the tourism sector is reshaping how travelers interact with digital services. This research introduces TourMate, a smart travel support platform that harnesses LLM capabilities to enhance the travel experience for both domestic and international tourists in Vietnam. The system features an AI-driven chatbot, personalized itinerary recommendations, and real-time travel insights, enabling seamless and adaptive assistance. Using natural language processing and machine learning, TourMate can understand user preferences, suggest optimized routes, and provide reliable local service recommendations. This study examines the implementation of LLMs within TourMate, assesses their impact on user engagement, and explores challenges such as data reliability, multilingual functionality, and responsiveness. The findings offer valuable insights into the development of AI-driven tourismapplications, contributing to the advancement of intelligent travel solutions. Research purpose: The purpose of this research is to investigate the application of Large Language Models (LLMs) in the tourism sector through the development of TourMate, an intelligent travel support platform. Specifically, the study aims to assess how LLM-driven features—such as personalized itinerary recommendations, adaptive chatbot support, and real-time travel insights—can enhance user engagement and improve the travel experience of domestic and international tourists in Vietnam. Research motivation: Tourism in Vietnam is experiencing rapid growth, with increasing demands for personalized, seamless, and digital-first travel services. However, existing solutions often lack adaptability, multilingual support, and contextual awareness. Recent advances in LLMs offer a promising opportunity to overcome these limitations by enabling intelligent, human-like interactions. This research is motivated by the need to bridge the gap between conventional travel platforms and the growing expectations of tech-savvy travelers, while also contributing to the digital transformation of the tourism industry. Research design, approach, and method: This study adopts a design science research approach, combining system design, prototyping, and user evaluation. The research process includes: (1) System Development – Designing and implementing the TourMate platform with key modules such as an LLM-based chatbot, itinerary optimizer, and real-time insight generator. (2) Experimental Evaluation – Conducting usability testing and user studies with both domestic and international tourists in Vietnam to assess engagement, satisfaction, and reliability. Main findings: The results indicate that the integration of LLMs significantly enhances user engagement by enabling more natural and context-aware interactions. TourMate was found effective in delivering personalized itineraries, providing accurate local recommendations, and supporting real-time decision-making. Nevertheless, challenges remain in terms of ensuring data reliability, maintaining fast response times, and addressing multilingual complexities. Practical/managerial implications: This research offers several implications for tourism stakeholders: - For service providers: LLM-based platforms can improve customer experience, increase loyalty, and reduce reliance on human support staff. - For destination managers: Intelligent insights can help optimize visitor flow, reduce congestion, and improve satisfaction. - For technology developers: The findings highlight the importance of balancing personalization with performance, and of designing scalable, multilingual AI systems. Overall, the study demonstrates that adopting LLM-driven solutions can accelerate the digital transformation of tourism, positioning Vietnam as a leader in smart tourism innovation.