Zero Hunger Network

Logo
Landing Page
Application Architecture

Inspiration

The idea for the Zero Hunger Network was inspired by the urgent need to address food insecurity and reduce food waste. Every year, millions of pounds of food go to waste, while many people struggle to find consistent access to nutritious food. We wanted to create a solution that bridges this gap by connecting those with surplus food to those in need, aiming to make a positive impact in our communities.

What it does

Zero Hunger Network is a platform that facilitates food donations by connecting donors, such as restaurants and grocery stores, with local food pantries and shelters. Using predictive AI, the platform analyzes donation patterns and forecasts future donation needs for each pantry. This allows us to provide real-time insights into where food is needed most, helping food pantries to better plan for and meet the demand.

How we built it

We built Zero Hunger Network using a mix of frontend and backend technologies. The backend was developed with Flask, integrating the Prophet forecasting tool to predict donation and distribution trends. Additionally, we integrated geolocation tools to match donors and recipients with nearby pantries. We used a database to store information about food pantries and historical donation data. The frontend was designed with vanilla HTML, CSS, and JavaScript.

Challenges we ran into

Initially, we used Firebase to host and serve our data and run predictions. However, we quickly found that the demand from running Prophet AI models made Firebase too costly to maintain for a project of this scale. Therefore, we had to migrate our back-end functions to Flask. This was an unforeseen circumstance that caused us to lose time.

Accomplishments that we're proud of

We are particularly proud of building a functional platform that can make a tangible difference in food distribution. Successfully implementing the predictive model for food donations was a significant milestone, as it enables us to forecast needs and reduce waste.

What we learned

Throughout this project, we learned predictive modeling and the importance of clean, structured data in making accurate forecasts. Working with location-based matching algorithms helped us appreciate the nuances of building solutions that scale geographically.

Chat GPT was used to assist in synthetic sample data generation