JournAI

landing page
web landing page
exploration options
ai preference questionair
scavenger hunt live tracking
adventure mode ai tour guide

Inspiration

As first-year college students, we remember how intimidating it felt to be at an unfamiliar campus. JournAI seeks to make exploring a college campus a fun and AI-enhanced experience.

What it does

JournAI is a web app that allows students to have an interactive, AI-enhanced experience with their school campus. There are two modes, scavenger hunt and adventure.

In scavenger hunt, a circle will appear on the map which contains a landmark which the player needs to find. The Google Gemini API generates riddles that when solved, will help the user find the landmark, as well as give interesting facts about the landmark. When the user gets close enough to the landmark, they win the scavenger hunt!

In adventure, Gemini API asks questions to the player to figure out their preferences and interests that will be used to personalize their tour with hints on how to reach the next destination on the map.

How we built it

JournAI is the culmination of over 30 hours of development and testing. Our front end consisted of Next.Js, React, Node.Js, Three.Js and Typescript. On our backend we utilized MongoDB Atlas, Google Cloud Instances, Google Map/Geolocation, Python, Gemini, Fetch.Ai and the FastAPI framework.

Challenges we ran into

The first challenge we ran into was correctly prompting the Gemini API for what we wanted. Since the riddles and hints rely on the response by Gemini, we needed to ensure that the response was in a certain format that we could use. The second challenge we ran into was tracking the player's location using the Geolocation API. It was difficult to make the tracking accurate and fast, so it is usable for the game winning conditions. The Final challenge we ran into use using the Fetch.AI queries with our backend in order to grab the Gemini prompts. There was an issue with the amount of supported imports on the blockchain. Pivoting to Cloud Instance we were able to utilize the Gemini API with our FastAPI python backend in order to deliver specialized user results.

Accomplishments that we're proud of

We are proud of the seamless integration of all of the technologies and frameworks we used to make one cohesive product. It was rewarding to see our idea come to life and create a tool that can genuinely benefit new students.

What we learned

We learned how powerful AI can be to elevate an experience. Integrating AI into the app made it more interactive and engaging for users, and we're excited to explore more possibilities in the future. We were also exposed to several new tech-stacks as well as company products with will allow us to develop apps faster and cleaner in the future.

What's next for JournAI

Moving forward, we plan to further develop the Gemini AI integration to provide even more engaging experiences for users. We also plan to add more landmarks, riddles, and personalized experiences to keep the app fresh and engaging. Lastly, we want to foster a community around the app, allowing users to share their experiences and challenges.