Left Overs

Inspiration

Food waste is a global crisis that, if solved, can create lasting benefits on numerous communities throughout the world. We have a love for cooking, however, noticed the large amounts of food left over in our fridge. According to the USDA, over 30 percent or 133 billion pounds of food are wasted in a year in US alone.

What it does

Our app asks the user to input different items they may have left over in their kitchen, and then returns a list of recipes. If the user selects a recipe, the instructions and further descriptions are then provided. Users are also prompted to sign in and favorite recipes are stored.

How we built it

We used a data set containing over 20,000 recipes on Epicurious.com, which had the food ingredients and directions, as well as ratings provided. The data set was then matched with a separate cleaned ingredients list. We then wrote Flask API. We used Google Clouds to deploy our Flask API on a gunicorn server as well as run the various scripts. This was to allow the recommendation processing of over millions of data points to be done not on the local device.

Challenges we ran into

The initial food ingredients categories, provided by the Epicurious data set contained a lot of noise, which required us to clean the data as well as find methods of gathering a cleaned ingredients list. Also, we had trouble setting the Google Clouds server due to lack of experience.

Accomplishments that we're proud of

We were able to work with different aspects of the development and come out with a completed project. Our algorithm works with decent accuracy and our app provides enough utility even for our own personal use. Setting up the sign in framework was quicker than anticipated.

What we learned

We learned a lot about creating servers, and how to create a back-end system and create an app using API calls. We all had to learn a bit of each other's area of expertise to increase working efficiency.