Inspiration

We were inspired mainly due to something that's becoming visible on social media, some food employees sharing videos of the massive amounts of untouched food that have to get thrown away at the end of the day. Seeing this perfectly fine food go to waste just stuck with us, making us wonder how many other restaurants might be doing the same. This led us to create a solution that can connect restaurants willing to give away leftover food with homeless shelters or food charities, either as a donation or for a flat fee,e for example 50. This way it can help reduce the massive amount of food that gets wasted by restaurants as well as helping provide food to the homeless, and food charity events.

What it does

Our project essentially connects Local Restaurants/Fast food chains with homeless shelters or food charity organizations. Restaurants can post how much food they have available in pounds, restaurants can choose if they want to charge or not for it. If restaurants don't charge it can be considered as a tax-deductible donation. Shelters/Food Charity organizations can browse restaurants near them that would fit their criteria, such as cuisine or the amount of food available. They can also use AI, which will act as a search query to help shelters/organizations find restaurants near them. On the restaurant side they can view the order and have the authority to accept or decline the order.

How we built it

We built this app using Next.js for the frontend and the backend of the framework We integrated Firebase in here, which manages the database for storing orders, chats, and authentication for users to log in. We also used tailwind.css for styling and Shadcn/ui to create a clean and accessible UI. To enhance the experience for the shelters/organizations, we added a chatbot using Gemini AI, which can help suggest restaurants to order from based on the shelters' preferences. Additionally, to handle payment, we added Stripe as our provider for a smooth checkout experience

Challenges we ran into

  • Stripe was initially hard to configure. We had issues following the documentation, but after looking at more examples that utilized Stripe, it made it easier.
  • Authentication was a bit challenging as we needed to set up secure routes for logged-in users and handle cases like logged-in sessions for the user when they navigate to different pages, because it would initially log them out, which took some time for debugging in the console to get it.
  • Making the UI really pop out, this just really took a lot of debugging and various reiterations to really get the UI right.

Accomplishments that we're proud of

  • We are proud of completing the most difficult features of our application. The AI, Agent Stripe, a brand new UI library, and many more were definitely tedious for us to complete. We are happy that we were able to complete our main key features, and we're also happy that we kept in a position where we can make an impact.

Another accomplishment that we found pleasing was that we were able to get a full-stack application to work while making it aesthetically pleasing.

What we learned

We learned that getting a full solution to a problem as big as this will take time; this isn't just a one-hackathon thing. Our team also learned how important LLMs are during this whole process, and how useful and stress-free they are to integrate within our code.

We learned that better time management is required, and a more structured approach to writing code is required to make efficient solutions. Furthermore, we learned how to use Tailwind CSS further, how to utilize and optimize Firebase (specifically the Firestore database) to ensure quick responses and accurate results.

Stripe also took a lot of time; last year, when we were at this hackathon, we just were not able to get Stripe working, and we pivoted to PayPal last year. This year, we are happy that we were actually able to get Stripe to work, and many of our team members found it to be a personal accomplishment as well.

What's next for Food Connect

  • Plan on hosting the website to initially get it kicked off and have its name out there, being spread
  • Making our chatbot more robust so it can handle even more complex queries
  • Implementing algorithm (like TikTok) suggestions based on the shelters' interests & preferences, this will display top matches at the top of the screen when the user goes to place an order.

Built With

Share this project:

Updates