Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Vaiage: A Multi-Agent Solution to Personalized Travel Planning

Binwen Liu, Jiexi Ge, Jiamin Wang

TL;DR

This work tackles dynamic, personalized travel planning by introducing Vaiage, a modular multi-agent system that combines LLM reasoning with real-time data via external APIs. A graph-based TravelGraph coordinates specialized agents (Chat, Information, Recommendation, Route, Strategy, Communication) to produce adaptive itineraries that respect user goals, constraints, and live conditions. In human-in-the-loop evaluations, Vaiage achieves an average of 8.5/10, outperforming baselines and demonstrating the value of coordinated agent reasoning and real-time information for feasible, personalized travel plans. The findings highlight the practical potential of integrating symbolic coordination with LLM-based planning for real-world decision tasks.

Abstract

Planning trips is a cognitively intensive task involving conflicting user preferences, dynamic external information, and multi-step temporal-spatial optimization. Traditional platforms often fall short - they provide static results, lack contextual adaptation, and fail to support real-time interaction or intent refinement. Our approach, Vaiage, addresses these challenges through a graph-structured multi-agent framework built around large language models (LLMs) that serve as both goal-conditioned recommenders and sequential planners. LLMs infer user intent, suggest personalized destinations and activities, and synthesize itineraries that align with contextual constraints such as budget, timing, group size, and weather. Through natural language interaction, structured tool use, and map-based feedback loops, Vaiage enables adaptive, explainable, and end-to-end travel planning grounded in both symbolic reasoning and conversational understanding. To evaluate Vaiage, we conducted human-in-the-loop experiments using rubric-based GPT-4 assessments and qualitative feedback. The full system achieved an average score of 8.5 out of 10, outperforming the no-strategy (7.2) and no-external-API (6.8) variants, particularly in feasibility. Qualitative analysis indicated that agent coordination - especially the Strategy and Information Agents - significantly improved itinerary quality by optimizing time use and integrating real-time context. These results demonstrate the effectiveness of combining LLM reasoning with symbolic agent coordination in open-ended, real-world planning tasks.

Vaiage: A Multi-Agent Solution to Personalized Travel Planning

TL;DR

This work tackles dynamic, personalized travel planning by introducing Vaiage, a modular multi-agent system that combines LLM reasoning with real-time data via external APIs. A graph-based TravelGraph coordinates specialized agents (Chat, Information, Recommendation, Route, Strategy, Communication) to produce adaptive itineraries that respect user goals, constraints, and live conditions. In human-in-the-loop evaluations, Vaiage achieves an average of 8.5/10, outperforming baselines and demonstrating the value of coordinated agent reasoning and real-time information for feasible, personalized travel plans. The findings highlight the practical potential of integrating symbolic coordination with LLM-based planning for real-world decision tasks.

Abstract

Planning trips is a cognitively intensive task involving conflicting user preferences, dynamic external information, and multi-step temporal-spatial optimization. Traditional platforms often fall short - they provide static results, lack contextual adaptation, and fail to support real-time interaction or intent refinement. Our approach, Vaiage, addresses these challenges through a graph-structured multi-agent framework built around large language models (LLMs) that serve as both goal-conditioned recommenders and sequential planners. LLMs infer user intent, suggest personalized destinations and activities, and synthesize itineraries that align with contextual constraints such as budget, timing, group size, and weather. Through natural language interaction, structured tool use, and map-based feedback loops, Vaiage enables adaptive, explainable, and end-to-end travel planning grounded in both symbolic reasoning and conversational understanding. To evaluate Vaiage, we conducted human-in-the-loop experiments using rubric-based GPT-4 assessments and qualitative feedback. The full system achieved an average score of 8.5 out of 10, outperforming the no-strategy (7.2) and no-external-API (6.8) variants, particularly in feasibility. Qualitative analysis indicated that agent coordination - especially the Strategy and Information Agents - significantly improved itinerary quality by optimizing time use and integrating real-time context. These results demonstrate the effectiveness of combining LLM reasoning with symbolic agent coordination in open-ended, real-world planning tasks.
Paper Structure (34 sections, 10 figures)

This paper contains 34 sections, 10 figures.

Table of Contents

  1. Introduction
  2. Background
  3. Temporal and Knowledge-Enhanced Reasoning with LLMs
  4. Personalized Recommendation with LLMs
  5. Methodology
  6. External Data Integration
  7. Multi-Agent Architecture
  8. LLM Reasoning and Prompt Design
  9. User Guides
  10. Getting Started
  11. Information Collection and Processing
  12. Attraction Selection
  13. Strategy Planning
  14. Route Planning
  15. Interactive Map Features
  16. ...and 19 more sections

Figures (10)

  • Figure 1: multi-agent workflow
  • Figure 2: Initial Chat Interface of Vaiage, showcasing the intuitive entry point for travel planning.
  • Figure 3: Information Collection and Processing, illustrating how Vaiage transforms user input into actionable travel insights.
  • Figure 4: Attraction Selection Interface: Left – Overlapping options for exploration; Right – Finalized attraction choices.
  • Figure 5: Strategy Planning Interface, highlighting car recommendations and trip adjustments for optimal scheduling.
  • ...and 5 more figures