Our inspiration for Appétit comes from our lack of creativity to come up with ideas for a meal. We find ourselves always eating out when we have perfectly good food and ingredients in our home. We would like to make it easier to remember what food items you have in your house as well as receive recipes ideas that promote healthy eating.

Also, in the US alone, we waste about 1.4 billion tons of food every year. [1] This statistic is stunning, and not in a good way. For that reason, we wanted to build an application that could help families and individuals reduce their food waste.

What it does

Appétit is a web application that simplifies everyday life by offering our users a way to alleviate the hassle of figuring out was to make for lunch or dinner and reduces the amount of eating out. With our application, you can keep track of the items in your fridge or cupboard, and receive personalized recipes with the ingredients you possess.

How we built it

Appétit was built with love and care, but mostly code. We have two components for this application. We have the front-end, which is the UI for the application which was built in Angular. The backend which hosts our API for the users and inventories was written in C# .Net 5. Several APIs were used in order to speed up our process. We use an API to read and decode UPC codes from products as well as an API to retrieve personalized recipes by ingredients. For detecting food using computer vision, we used tensorflow.js with the cocoSSD model.

Challenges we ran into

We run into quite a lot of challenges when came the time to put all of the components together. Things didn't go as smoothly as expected, and of course, lots of bugs occurred. Despite that, we powered through and managed to produce a working prototype. Some of the APIs were a little too restrictive, therefore we opted for web scraping in order to increase the amount of requests we could make. Finally, we had some deployment issues where some configurations were not taking effect in production.

Accomplishments that we're proud of

We are proud to have been able to produce a working prototype in the 36 hours of the hackathon. We have been able to accomplish quite a lot in that time, such as item recognition through AI, barcoding scanning, user and inventories databases, recipe suggestions and more...

What we learned

Throughout this project, we quickly developed new skills. We learned that projects are more easily made with friends which share the same ideologies as you. These experiences provided us with an appreciation for the development process and research that goes behind creating a new product for the world to see.

What's next for Appétit

Have a lot of other great ideas for Appétit! This only being a prototype, we would like to improve on the tracking of items in your inventories. Find ways to make it less tedious to input items in them. We would like to improve our recipe suggestions to take into account the time of day, allergies and preferences. We would also like to train our own TensorFlow model to have more flexibility on the items we want to be able to detect and have more control over the accuracy.

[1]: Food and Agriculture Organization (FAO) of the United Nations (UN)

Built With

Share this project: