Inspiration
Recently, as we all struggled through spring break trials and tribulations, we all shared a common enemy: our own friends. As we tried to plan out where we would go, what we would do, and how we would organize everything, we lost countless hours sending texts into the void, never to be answered. In order to reduce our own future pain, as well as the pain of anyone else trying to plan a group trip, we created Triplet - a group brainstorming and itinerary generating application.
What it does
Triplet takes a group of users, gathers their ideas, then uses AI to collate all of the ideas into a set of potential activities. The users then vote on the activities, swiping directionally (like Tinder!) to indicate if they liked an idea or not.
How we built it
We built our application using Next.js as our framework, and Firebase as our hosting solution and database solution. For styling, we used tailwind, and for AI generation we used GPT 4.0.
Challenges we ran into
Our very first challenge was one of the most common stories that you hear at hackathons. Triplet was our second idea. Our initial plan was to create a AR tour guide app that would allow you to explore foreign cities following a AR trail laid out in front of you, with waypoints in the sky marking notable landmarks along the tour. However, after working for a couple hours, we ran into an issue. All of the solutions that we tried were either not maintained, or implemented so complexly that we would not be able to build our own application on top of their platform. We also explored building our own AR solution, but Apple repeatedly stopped us through their confusing development tools and their misguided compass. This led to our second idea, Triplet, having limited development time. We had lost 8 hours, meaning that with sleeping time considered, we only had 10 hours left for our application.
Another challenge we faced was different experience levels with our chosen frameworks and languages. Different members had varying levels of experience with the technologies in our stack, and due to that, we only had varying levels of work that each member was able to do. Also, with limited time, it was difficult to justify slowing down and teaching the framework to all of our members, as we needed to focus on pushing our development forward.
Accomplishments that we're proud of
We are incredibly proud of our resilience, as pivoting ideas at such a late hour in the day hit our morale, and made working on our idea that much more difficult.
What we learned
We definitely learned that there should be a faster judgment call as to when to continue or bail on an idea, as if we had been able to realize the problems with our initial idea sooner, and pivot sooner, we would have had more time to implement Triplet, leading to a more polished and fleshed out idea.
What's next for Triplet
We envision Triplet as a full assistant, being able to book tickets and purchase housing for our users, as well as serving as a decision making tool for all of their planning needs, such as exact housing location and nightly dinner plans. We also want to track user preferences over time, so we can use AI to generate activities and vacation ideas that they are more likely to select.
Log in or sign up for Devpost to join the conversation.