• 

  Logo for my application: SaveAServing

What it does

This application helps business owners predict future food orders using ML, enabling them to plan and manage meal preparation at an optimal level to reduce waste. Additionally, the application connects businesses, donation centers, and volunteers, allowing businesses to donate surplus food, and volunteers to help deliver it to donation centers.

How I built it

I developed a web application using React JS, Supabase (database), Day.js, and AWS SES. For the machine learning component, I used Python, FastAPI, and Prophet to train a model for order prediction.

Challenges I ran into

Initially, I tried using a different ML model, but its mean squared error (MSE) was too high for my use case. I then switched to the Prophet model, which had a much lower MSE and significantly improved prediction accuracy.

Accomplishments that I am proud of

This is my first application that incorporates a machine learning model, and I’m proud to have learned and implemented it successfully. I'm also proud that this application helps bridge the gap between restaurants and people in need, reducing food waste and combating hunger.

What I learned

I learned how to train and implement a machine learning model. I also learned to use AWS SES for sending emails. More than anything, I learned that technology holds real power to address important societal challenges.

What's next for SaveAServing

Since this is a time-sensitive and hyper-local application, I plan to implement faster communication methods such as text messages or WhatsApp notifications. I also aim to integrate location-based features to match businesses and donation centers with nearby volunteers. In the future, I’d like to introduce a points system to recognize and reward the contributions of restaurants and volunteers. Additionally, I plan to add team and gamification features to boost engagement.

Built With

• aws-ses
• day.js
• javascript
• mui
• prophet
• python
• react
• react-router
• supabase