FlyWise

Inspiration

Travel planning is stressful — flight prices fluctuate constantly, deals vanish in hours, and comparing dozens of hotel tabs is exhausting. We wanted to make the process smarter, simpler, and more personalized. FlyWise was born from the idea of using AI to remove the guesswork and help travelers make confident, data-driven decisions about when to book, where to stay, and how to prepare for their trips.

What it does

FlyWise is an AI-powered travel assistant that predicts the best time to book flights, recommends personalized hotels, and even provides quick language tips for travelers visiting new destinations. The platform analyzes live flight data, budget preferences, and trip purposes to deliver tailored recommendations. Once users find their ideal plan, FlyWise connects directly to Booking.com for real-time booking options — creating a smooth end-to-end travel experience.

How we built it

We built the frontend with React + Vite for a fast and interactive interface, and the backend with FastAPI (Python) to handle AI endpoints and data logic. We used the Google Gemini Flash Latest model for generating natural-language hotel insights and travel tips, while Snowflake API handled data structuring and queries. FlightAPI.io supplied real-time flight pricing and route data. Finally, we unified the frontend and backend into one domain using FastAPI’s static serving and deployed it via Render for a single-link experience.

Challenges we ran into

We faced issues with cleaning and parsing Gemini’s generative JSON output, securing environment variables, and merging multiple services (React, FastAPI, Gemini, Snowflake) into one deployment. Another challenge was ensuring low-latency communication between APIs while maintaining accurate, context-aware AI responses.

Accomplishments that we're proud of

We successfully created a fully functional AI travel assistant that integrates live flight data, generative AI recommendations, and booking automation. We built a unified platform deployable under one link, developed a clean and user-friendly UI, and incorporated Gemini for both hotel suggestions and language travel guidance.

What we learned

We learned how to integrate multiple AI and data services securely, manage real-time API responses, and use prompt engineering to achieve structured, relevant Gemini outputs. The project also strengthened our collaboration and problem-solving skills under time pressure, uniting backend, AI, and frontend systems seamlessly.

What's next for FlyWise

Next, we plan to integrate real-time hotel availability and multi-language AI chat support for instant translation or local advice. We’ll also enhance budget optimization dashboards, add user profiles for saving favorite trips, and expand our recommendation system to include dining, cultural experiences, and itinerary generation — all powered by Gemini and live data APIs.