A friend of one us created a food delivery app that uses students specifically to deliver food to other students. We liked the idea of giving students another chance to make some cash going throughout their busy day. One of us has also overheard a fellow classmate at FIU describe her daily trip to FIU by bus and were shocked that it took her around an hour and 30 minutes just to get to school, most of that time being spent waiting for the bus or being stuck in traffic. Both inspired us to create an app that solves two issues: a lack of jobs that cater to busy schedules and a lack of travel options for commuting students.
Our app creates an easy-to-use platform for commuting students to find rides from other students based on their route to school or other school-related events they may be planning to join. It allows students to post both as a rider or driver and best matches students based on the how much their routes overlap. It also uses zip codes to find which students are nearby and available to provide transportation or are looking for transportation. You can select what times you're looking for a ride and can even upload your schedule directly to match with students who take the same class or have a similar schedule. This makes it easy to not only set up one ride for the day, but set up reoccurring trips with the same driver weekly once a route and schedule match has been found.
We used Google AI Studio with Google Gemini API to make the foundation of our website and add any extra tabs or features that we thought of along the way. We would then test its functionality and redefine the prompt until the feature being added functioned satisfactorily.
A challenge we ran into was implementing Auth0 or identification to restrict access of the app to only students, as they are our target audience. This was one of the most important aspects of our app, because if it was open to the public, that would defeat the purpose and destroy the trustworthiness of making an app for students. Another challenge we faced was getting the map to show accurate times for the routes it was producing. When putting in a route for which we knew the time it would take to travel, it would give a time far lower than that time.
We managed to include a way for students to upload their class schedules directly rather than typing it in manually. It uses the same method google calendar uses to upload the canvas assignment calendar or the class schedule that can be downloaded from MyFIU by accepting .ics files. This allows riders and drivers to match based not only on routes, but on schedules too. Another accomplishment we made was the chat system between both potential parties. We made a way to connect directly and make offers similar to a user-run store platform like Facebook Marketplace. The rider has the ability to accept an offer and choose what payment method they prefer, and the driver can choose whether or not to accept said offer.
We learned that it is very challenging and time-consuming to build an app or website, even with the use of AI. We also learned how to use agents to facilitate the data collection and functions of our app that needed external influence.
The next step would be to iron out the details of the app, clean up the code, improve the accuracy of features like the map and the mileage rates, and add any small features we deem necessary for the convenience of the user. After that, the next step would be marketing the app to students at FIU and possibly other campuses in South Florida and receive feedback before pushing out the app further.
Log in or sign up for Devpost to join the conversation.