As we were brainstorming together on the first day of the hackathon we very quickly came to the conclusion that we would draw heavily from the Tinder UI as we liked the simplicity and clean feel of Tinder. However, the direction we were looking to take with the UI was very different from Tinder, focused on clothing rather than relationships. One of our team members Justine, felt as though when shopping, there are plenty of items to choose from, but trawling through them and trying to find what you prefer is an arduous task. We wanted to create an app that would tag items of clothing with certain tags such as color, brand, fit, size, etc. and as you swipe left and right the app tailors what it shows to match what you frequently like. We wanted the final product to learn and then display a users style to them from a variety of sellers.
What it does
By clicking like and dislike on items of clothing, the app learns what tags you prefer over others, and shows clothing items that align closely with what you frequently like. Therefore over time, the app will learn what your "style" is and then trawl through websites and stores to find what you are looking for, thereby removing the need for you to dedicate time to find clothes that fit your aesthetic.
How we built it
We initially started by breaking up the project into two parts, with half of the team dedicated to each. One half would be working on Android SDK primarily focused on the frontend and secondarily on the backend, while the other would be working in Eclipse with Java, centering on the backend, primarily the system that would try to learn the users' preferences and determine what tags are associated with what articles of clothing. However, after speaking with several members of the UL team here we decided that doing all of the work in Android Studio was the way to go. We then split into two halves again, one dedicated to researching and developing the frontend and UI while the other half was focused on the backend.
Challenges we ran into
There were a few problems that we ran into, the foremost being our inexperience with hackathons, as this was the first hackathon any of us had attended, so getting into the right rhythm took a little bit of time. The first problem we faced was fairly early on, and that was the issue of using Android SDK, as we had no prior experience and had very little idea as to how the libraries worked. We overcame this via generous Googling and stack overflow use in order to understand what we needed to know to reach our goal. The second issue followed fairly quickly after, wherein we had no idea how to mesh together our work in Eclipse and Android SDK. It was only on the second day that we solved this problem by circumventing it entirely by switching exclusively to Android Studio. Unfortunately, as we had no place to stay, we were relegated to sleeping on the floor, which did not exactly boost our morale in the morning and we had to struggle to stay motivated and on task for a few hours. Thankfully we got back on our tempo come noon and it carried us well through the rest of the day.
Accomplishments that we're proud of
By and large turning in anything at all was our greatest accomplishment, as we had no experience with Android Studio, and we ended up working solely through it. Learning and implementing everything over the course of two and a half days greatly limited our speed, and while we had to cut a great deal of our original vision, we are proud of being able to submit a functioning app with just what we learned in two and a half days at our first hackathon. Our team came in with two members with an intermediate understanding of Java, and two beginners, and by the end of the event each of us was able to learn on our own and teach each other, so as to maximize our ability to take in and process new information.
What we learned
Our group had zero prior experience in Android SDK so almost all of our project was built off of what we could Google. One of our members also happened to delete one of our xml files relatively early on quickly teaching us to backup anything and everything even remotely related to our work. We came in with almost no idea what we were doing, or what we wanted to accomplish, and we ended with a project leagues above what we had believed we were capable of when we first set out to learn about Android Studio. We learned the value of preserverance as well, when we were feeling defeated and undoubtedly cranky on Saturday morning we chose to power through and slog through our work until we wore off the morning drowsiness.
What's next for That's Sew You
We want to continue to refine our UI to make it more polished, still drawing on Tinder's influence for the final product. And we also want to move to a server so that clothing doesn't have to be manually input as it was for this project and so that we can draw from many different stores databases of clothing and perhaps even accesories should we continue to expand our software. We also would like to actually implement our idea of the app learning from the users' preferences to provide a tailored service, as we were unable to do so within the time constraints, and with our limited experience with Android SDK.