Munch

Approachable way to start a healthy lifestyle during the pandemic.

Comment

Inspiration

During this pandemic, we encountered a problem common among young adults: we didn’t know how to cook and were stuck with unhealthy food options at home. Many students like us do not have access to healthy food during this pandemic and many working parents do not have the time to prepare healthy recipes. Learning to cook is daunting because new learners are faced with a jumble of unclear cooking articles and long videos that are difficult to navigate and contain extra information. Someone can spend 5 minutes reading a recipe only to notice they didn’t have a key ingredient. This makes finding recipes very inefficient and time-consuming.

What it does

To solve this issue, we created Munch, a social media platform designed to share concise and healthy recipes. Munch allows people to quickly search for recipes by specifying ingredients, tags, or titles. The recipes cut to the chase and only display important information with minimal clutter. Each step of the recipe will come with a video that demonstrates what to do. Another goal is to implement an algorithm that helps people find recipes they love based on their feedback on previous recipes. Users can create and share their own recipes by posting the necessary info. They Include the contents and we save it to our database. As a result, the end product provides an efficient and intuitive way for people to learn to cook as well as a universal platform for users to post and find recipes that will be tailored to them.

How we built it

We built this using the MERN javascript stack. We also used TensorFlow for the machine learning section. For the actual app, MERN was ideal since it allowed us to quickly share the code between our teammates and edit the components separately. React is very modular so that allows us to put the app together like pieces of a puzzle. Using MongoDB was great since it is extremely quick and has great integration with both React and Javascript. It's incredibly fast and efficient and allows us to store data however we want it to.

Challenges we ran into

The search feature was a challenging portion of the app since it required strong algorithmic knowledge. We had to figure out how exactly to make it work and make sure that we included all the data we intended while not resulting in false positives. It had many bugs that we had to sort through.

Accomplishments that we're proud of

Our search feature and overall design are great. The search feature is especially nice since the search query can be very vague and the program still outputs the closest match. It doesn't need an exact match at all. Rather, it just has to be close enough. This will be further optimized with machine learning. The overall design is also very scalable. In all the components we made an emphasis on writing clean, scalable, modular code that would be able to change to whatever we need it to. Our organization is very clean and our code is very systematic.

What we learned

We learned a lot in algorithmic knowledge and strengthened our knowledge of the MERN stack. In this project, we had to write many React components and thus we got a lot of practice dealing with Parent/Child relations and state functions and more.

What's next for Munch

We have written pseudo code to start to implement machine learning into the project and finding ways to recommend recipes for users is the next step. Beyond that, we would like to port to mobile and get users on the app to get feedback.

Share this project:

Updates