Eventure: Your AI-Powered Travel Assistant
Inspiration
Our inspiration came from a real frustration we had, where one of our teammates was going to New York but struggled to make a proper itinerary while compensating for her and her family's specific interests and dietary restrictions, since they've never been to New York.
That's why we created Eventure to be able to eliminate the struggle in travel planning and just regular day trips in general, by being able to input in every detail about your trip and getting a fully planned itinerary for anywhere you need.
What it Does
Eventure takes in a list of preferences and details about your trip and outputs either a single-day or multi-day itinerary.
Users have complete control over a wide range of criteria, including:
- Dates & Times: Define the exact length of your trip.
- Group Size: Plan for solo adventures or group getaways.
- Budget: Set per-activity cost expectations.
- Location & Radius: Choose a central point and how far you're willing to travel.
- Activity Preferences: Select from categories like Restaurants, Museums, Nature, Shopping, and more.
- Dietary Needs: Ensure every meal recommendation fits your requirements.
- Accessibility needs: Whether it's wheelchair friendly doors, elevators, etc.
With these filters, our application constructs a query for the Google Gemini API, which then generates a complete, day-by-day (or singular day) itinerary designed just for your needs.
How We Built It (Our Tech Stack)
We chose a scalable tech stack to bring Eventure to life, leveraging the power of the JavaScript ecosystem and Google Cloud.
Design & Planning:
- Figma: Our entire multi-page design, from wireframes to high-fidelity mockups, was crafted in Figma to create a clear vision for the UI/UX.
Frontend:
- Next.js & React
- TypeScript
- Tailwind CSS
Backend & Google Cloud
- Express.js
- Google Cloud Database: To manage user data, sessions, and saved itineraries.
- Google Cloud Console & APIs: We deeply integrated two key services:
- Google Places API: To power our intelligent location search and provide rich data on points of interest.
- Google Calendar API: To share and export the itinerary to google calendar for better viewing
- Google Gemini API: The core AI engine that generates the custom itineraries.
Challenges We Ran Into
Mastering the AI Prompt: Crafting the perfect prompt for the Gemini API was a pretty big challenge. Initial queries were too broad and returned generic data. We overcame this through carefully structuring our JSON-style query to be more descriptive and constrained, which dramatically improved the quality and relevance of the generated itineraries.
Outputting ProperIy: Our BIGGEST challenge involved getting it to actually show on the page after taking in the input, the gemini prompt output is a very difficult thing to get down and especially allowing it to appear on the frontend with the specific output since it has to match perfectly and since the frontend and backend developed it separately it was difficult to match up the outputs perfectly. We had the same issue with allowing it to appear on Google Calendar as well, due to it being a very specific type of output and it being different from how we output it onto the frontend.
Accomplishments We're Proud Of
In the given timeframe, we achieved several milestones that we're incredibly proud of.
A Seamless and Intuitive UI: We successfully translated our complete, multi-functional Figma design into a fully interactive and responsive user experience that guides users effortlessly through the planning process.
Harnessing the Google Ecosystem: More than just using a single API, we deeply integrated a suite of Google services (Places, Gemini, Cloud Database) to work in concert, creating a powerful and intelligent application. As well as using the Google OAuth 2.0 to allow you to log in with Google, show your name as it appears on your gmail and use it to allow us to edit your Google Calendar with the outputted itinerary. This showcases our ability to work within a major cloud provider's ecosystem.
What We Learned
This project was a massive learning experience as none of us had really worked too much using Google's services/APIS so it was amazing seeing how much we were able to implement especially in such a short timeframe especially with such a clean UI.
What's Next for Eventure
Our vision for Eventure extends far beyond a personal planning tool since it can work with both your daily life as well as your special moments. Our next step includes:
- Fully implementing our planned features and cleaning up the ones already implemented.
- Clean up the Gemini Prompts to allow for more seamless and better-suited itineraries.
Built With
- express.js
- gemini
- google-database
- google-places
- json
- next.js
- node.js
- react
- typescript

Log in or sign up for Devpost to join the conversation.