During the pandemic many are stuck at home with little to do but learn cook. While this is fine for many, people with dietary restrictions find themselves having to sift through recipes to find one that fits their needs. Our website changes that by aggregating these recipes and filtering them to what our users want. After creating and tasting the meal, our users can also share and rate the recipe with their friends!
What it does
We build a website that takes Dietary Restrictions, Allergies, Pre-existing Conditions and Family History to give food recommendations based on the choices available as well as how likely the person is to get a certain health condition.
Once signed up the user will get food recommendation and its recipe as well as the info on the ingredients.
How we built it
We used React and firebase for the frontend and login authentication and then are using an API which gives food recommendation based on the search.
Challenges we ran into
When initially developing the project, we wanted to include an ML model but we weren't able to find the best dataset of function to achieve >90 accuracy level for the model. Hence, the machine learning algorithm sat between 50-70% model accuracy. After seeing the results we decided that our time was better spent on the website however, even then we had to deal with many strange (but not uncommon) issues such as CSS acting funny, git not allowing pull requests, and general internet issues.
Accomplishments that we're proud of
Collaborating and communication remotely was definitely something we are proud of because many simple tasks become longer and with the all the scheduling and troubleshooting we had to do, being able to create a hackathon project under time constraints could be considered an accomplishment.
What we learned
New methods of solving problems using React and python in addition to understanding the scope of the project. Having many different specialties and perspectives allowed us to view the project calmly especially as the deadline loomed over us.
What's next for Safe Menu
Full backend integration with an ML model that would allow us to predict the user's food preferences and their chances of developing an illness and general UI fixes.