Wandere.AI
AI Meets Wanderlust: Personalized travel plans tailored to your preferences, location, and activities. Effortless trips, anytime, anywhere.
Created on 9th November 2024
•
Wandere.AI
AI Meets Wanderlust: Personalized travel plans tailored to your preferences, location, and activities. Effortless trips, anytime, anywhere.
The problem Wandere.AI solves
Travel planning can be overwhelming. We frequently spend hours researching destinations, comparing prices, and creating itineraries that suit their budget and preferences. Moreover, getting local insights and historical context usually necessitates hiring guides, which can be expensive and inconvenient.
Our app simplifies this entire process by generating personalised, budget-friendly itineraries in minutes, along with cultural and historical insights that would normally require a local guide. By offering affordable, AI-driven itineraries, we allow travelers to explore their destination like a local—saving both time and money, while enriching their travel experience with meaningful information.
Each plan is customized, up-to-date, and available as a downloadable PDF or sent directly to your email. With accurate information and tailored recommendations, it’s the perfect tool for travelers wanting a stress-free adventure!
Challenges we ran into
1.Accurate API Integration for Itinerary Generation
We initially struggled with integrating AI APIs to generate detailed itineraries, particularly in formatting responses for different travel days and preferences. After trying various prompts and API configurations, we refined our request structure, ensuring consistency and high-quality results.
2.PDF Formatting for Itinerary
Presenting information in a readable PDF format was another challenge. Early versions didn’t maintain alignment, and important details were often missed. We experimented with different libraries and formats, finally optimizing the layout for a clean, professional look.
3.We wanted to add maps for showing routes, travel times, and locations but found the management of location data a challenge, as well as real-time accuracy with open-source maps proved to be difficult. This is definitely something we will improve in later updates
4.It was difficult to maintain the usage of different AI models and create prompts or wordings that suited specific sections of the itinerary. Testing and refitting the prompts helped us to finally succeed in upgrading the quality of the output while experimenting with models that best suited specific tasks.
- Local guides: we provided the feature to the app for suggesting local places and attractions, and the users were seeing suggestions locally. There were issues in being fully integrated into the itinerary generation flow. Although the feature is workable in isolation, technical and compatibility issues didn't enable us to mount it within the itinerary interface during the hackathon.
Tracks Applied (4)
Best Use of MongoDB Atlas
Major League Hacking
Best Use of Auth0
Major League Hacking
Best use of GitHub
GitHub Education