AI-Based Travel Planning System Using Large Language Models
Keywords:
Artificial Intelligence, Budget constraint, Groq API, Large language models, Prompt engineering, Travel planning systems, Web application architectureAbstract
Travel planning is a multi-constraint problem that involves balancing budget, duration, transportation, accommodation, and personal preferences. Existing online travel platforms largely focus on booking individual services and often fail to generate cohesive, personalized itineraries that strictly adhere to user-defined financial limits. This paper presents the design and implementation of an AI-based travel planning system that generates realistic, budget-constrained travel itineraries using Large Language Models (LLMs). The proposed system adopts a modular three-layer architecture consisting of a web-based frontend, a backend orchestration layer, and an AI inference layer hosted on Groq. User inputs such as destination, total budget, number of days, and travel preferences are transformed into structured prompts using strict prompt engineering techniques. Unlike conventional approaches where budget is treated as contextual information, this system enforces budget as a hard constraint, ensuring that the generated itineraries do not exceed the specified financial limit. The system outputs a structured itinerary comprising a budget breakdown, day-wise travel plan, and practical local tips, formatted in controlled HTML for consistent presentation. Additionally, transport booking options are abstracted at the user interface level to avoid dependency on region-specific service providers, enhancing scalability and geographic neutrality. Experimental observations indicate that explicit prompt constraints significantly improve itinerary realism, cost adherence, and output consistency. The results demonstrate that careful prompt design combined with modular system architecture enables Large Language Models to function effectively in real-world, constraint-driven applications such as travel planning.
References
N. J. Nilsson, Artificial intelligence: a new synthesis. San Francisco, Calif: Kaufmann, 1998.
T. M. Mitchell, “Does Machine Learning Really Work?” AI Magazine, vol. 18, no. 3, pp. 11–11, Sep. 1997.
E. Alpaydın, “Introduction to Machine Learning,” MIT Press, Mar. 17, 2020.
C. C. Aggarwal, “An Introduction to Recommender Systems,” Recommender Systems, pp. 1–28, 2016.
C. C. Aggarwal, Recommender Systems. Cham: Springer International Publishing, 2016.
F. Ricci, L. Rokach, and B. Shapira, “Recommender Systems: Techniques, Applications, and Challenges,” Recommender Systems Handbook, pp. 1–35, Nov. 2021.
E. Turban, “Decision Support and Business Intelligence System,” Pearson Education, 2025.
T. B. Brown, B. Mann, N. Ryder, “Language Models Are Few-Shot Learners,” Arxiv.Org, vol. 4, no. 33, May 2020.
J. Wei, X. Wang, “Chain of Thought Prompting Elicits Reasoning in Large Language Models,” arXiv:2201.11903, vol. 1, Oct. 2022.
P. Liu, W. Yuan, J. Fu, Z. Jiang, H. Hayashi, and G. Neubig, “Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing,” ACM Computing Surveys, vol. 55, no. 9, Sep. 2022.
I. Sommerville, Software engineering, 10th ed. Harlow: Pearson Education, 2015.
R. Pressman, “Software Engineering: A Practitioner’s Approach, 17th Ed.” 2005.
M. Fowler and D. Rice, “Patterns of Enterprise Application Architecture,” Martinfowler.com, 2012.
E. Gamma and R. Helm, “Design Patterns: Elements of Reusable Object-Oriented Software,” O’Reilly | Safari, 2019.
B. Shneiderman and C. Plaisant, Designing the user interface: strategies for effective human-computer interaction. Boston: Addison-Wesley, 2010.
B. Shneiderman, C. Plaisant, M. Cohen, S. Jacobs, N. Elmqvist, and N. Diakopoulos, “Grand challenges for HCI researchers,” Interactions, vol. 23, no. 5, pp. 24–25, Aug. 2016.
L. Floridi, The Ethics of Artificial Intelligence. Oxford University Press, 2023.
