Inspiration
Let’s face it: being broke college students isn’t easy. We want to impress our dates with fancy dinners, but we just don’t have the money for an air fryer. Or maybe we need an iClicker for that last-minute Gen Chem class, but spending $50 on something we’ll use once is just not in the budget. As a result, we turn to unsustainable and ethically questionable companies like Temu and Shein. We buy cheap, low-quality items, use them once, and throw them out. A 2021 survey by Ladder and OnePoll found that Americans, on average, spend $193.08 each month on online shopping for non-essentials and impulse buys. This totals $2,317 annually and nearly $200,000 over a lifetime. That’s where Appli comes in. We built this platform for people just like us—college students and part-time workers who need access to items in a sustainable way but don’t want to break the bank. Whether it's borrowing a blender for that smoothie to live healthier or a bike for a spontaneous weekend getaway, Appli lets users borrow everyday items for free. Appli’s unique point system is based on reciprocity: the more you lend out, the more you earn and can borrow. We’re a community of people helping each other out. With profile verification and photo documentation, we’ve created a trustworthy, secure place where you can borrow what you need without the awkwardness or risk. As broke college students, we’re all trying to make our tight budgets stretch, and Appli makes life a little easier and more sustainable. We’re one step closer to a greener world where people spend less and share more.
What it does
Our application not only facilitates the sharing of resources, but its AI recommendations allow them to get what they truly need. Upon opening the app, users see a for-you page with items generated based on what is currently available. They can scroll through this page to find items they are looking to borrow and add the items they are seeking to lend out. Moreover, the other pages of the application allow them to view potential messages with venders and buyers and their profile with their number of points.
How we built it
We decided to use React-Native for the frontend due to its design capabilities and Firebase for the database because of its real-time data handling. To retrieve information from Firebase, we used an intermediary backend platform called Flask that could receive this retrieval request to the API through HTTP and send information back to the client--a process that involves both the backend and frontend. The information we then gathered was fed to the AI models we used (Groq and LLaMA) and used to generate tags and recommendations for image items in the backend. On the frontend side, we implemented various methods in React-Native to allow for scrolling in the app and aesthetic color themes.
Challenges we ran into
Having little experience in integrating the various parts of a tech stack, particularly frontend, backend, and databases, before this hackathon, one of the hurdles we had to overcome was the learning curve for the platforms React-Native, Firebase, and Flask. In particular, we had trouble communicating between React-Native and Firebase and vice versa. However, we discovered the intermediary through Flask, which we used for the backend, and Ngrok, which allowed us to directly communicate between the backend and frontend. This process enabled us to do direct testing on multiple devices and mimic how a real application would act in communicating with a remote server.
Accomplishments that we're proud of
As our first time programming an application, we are proud of how much we learned and the progress able to make to implement our plans. Indeed, what we developed for this hackathon serves as the draft for a web application that could help both society and the environment.
What we learned
We went through multiple trial-and-error debugging sessions, from trying to synchronize the state between client and server to discovering when to use POST vs. GET. Some of the errors we found in debugging led us to learn that sometimes the largest impacts are caused by the smallest and least expected mistakes. By the end of the hackathon, we were therefore not just able to learn more about the skills used in developing a web application but also gain a deep understanding of problem-solving beyond the classroom and in industry--where teams work together to meet their goals.
What's next for Appli
As our next steps, we hope to fully implement and expand on the AI recommendations given to the user based on their borrowing history. We also want to bring multiple users to the platform and allow them to message each other and see their accumulation of points based on what they lend out.
Log in or sign up for Devpost to join the conversation.