Sidequest

Sidequest was inspired by the idea that the most memorable parts of travel often come from spontaneous adventures, the sidequests. Instead of rigid itineraries, we wanted to create a platform that brings spontaneity and social connection back into travel planning.

Sidequest is a web application that uses AI to generate personalized travel itineraries based on a user’s interests, destination, and travel dates. It also matches travelers with others planning similar trips by comparing user profiles and trip overlap. The app recommends local events, hidden gems, and unique experience, turning a basic itinerary into an adventure shared with like-minded people.

We built the frontend using React.js and Tailwind CSS to create a responsive interface. The backend runs on Flask, handling itinerary generation and user matching logic. We used Gemini for natural language-based itinerary generation and to compare user interests via embeddings. MongoDB stores user profiles and trip data, and we integrated the Google Places and Ticketmaster APIs to source real-world locations and events.

Our main challenges included defining what makes two travelers compatible, integrating structured and unstructured data across APIs, and aligning AI-generated content with real-world activities. Building a cohesive product across frontend, backend, and AI components under the pressure of a hackathon timeline was also a major hurdle.

We’re proud to have built a full-stack, AI-powered app in under 36 hours. We developed a working traveler-matching algorithm, designed a polished user experience, and brought together multiple technologies to deliver something both functional and fun.

We learned how to quickly integrate and align multiple APIs, how to use embeddings for user similarity, and how to use integrate databases. This project also pushed us to think critically about prompt design and data representation in AI-driven applications.

Moving forward, we plan to add messaging and group trip planning features, expand destination and event coverage, and build a mobile version for real-time, on-the-go planning. We also want to refine the AI to support collaborative itinerary edits and adapt suggestions based on user feedback.