CulturAI

Inspiration

Traveling should be about discovering new places, not juggling countless apps. We noticed how trip planning often means switching between platforms for activities, hotels, translators, and budgeting — a process that takes away the joy of exploration. That’s where the idea for CulturAI was born — an all-in-one travel companion that makes exploring the world effortless, personalized, and connected.

What it does

CulturAI provides personalized activity recommendations based on the user's interests and destination, next-destination suggestions tailored to user's travel style and budget, and a built-in language translator to bridge communication barriers.

How we built it

On the frontend, we designed an intuitive and responsive interface using React components and Tailwind CSS, ensuring users can easily explore activities, destinations, and translations in one place. Each feature interacts directly with the backend through dynamic API calls, displaying AI-generated insights in real time. Our backend, built with FastAPI, handles all the logic and communication with Gemini. It has modular endpoints for activity recommendations, destination suggestions, translations, user profiles, and a conversational chat assistant. We used Pydantic models to validate and structure data consistently across all endpoints, and CORS middleware to ensure secure communication between the frontend and backend. The Google Gemini API is the heart of our intelligence layer . It powers the reasoning behind recommendations and translations. We carefully designed prompts to guide Gemini into producing structured and relevant results that can be parsed and displayed cleanly on the frontend.

Challenges we ran into

Some of our challenges include integrating multiple APIs (Gemini + custom endpoints) while maintaining consistent data flow, parsing Gemini’s responses into structured JSON outputs, and also designing an UI that displays complex recommendations clearly and attractively.

Accomplishments that we're proud of

First of all, we successfully built a fully functional AI-driven travel app combining activity, destination, and translation services within 24 hours! Also, we successfully integrated Google Gemini for intelligent, context-aware recommendations. Our front end was another thing that we were particularly proud of. It looks pretty, organized, and easy to navigate for users.

What we learned

We learned how to connect LLM APIs (like Gemini) effectively to structured user-facing apps, the power of Next.js and FastAPI for building scalable, production-ready prototypes quickly, and the value of good UX in making AI-powered results understandable. Most importantly, we learned how to collaborate effectively as a team under time pressure.

What's next for CulturAI

For our next steps, we will implement flight & accomodations search functionality, budget optimizer, speech-to-speech translator that handle spoken language in real time, an interactive travel map that generate personalized travel routes, login and signup functionality so that the users can make an account and save their travel plans