Inspiration

Recently we've all started to notice a trend at restaurants. When you get there, one of the biggest challenges is what do we order? And there are a few ways people try to tackle this:

  • Ask for recommendations of food
  • Text their foodie friend for help
  • Start browsing yelp and reading reviews to try to grasp what is the best thing to order

However, sometimes recommendations by other people don't necessarily fit our taste, which can lead to a possibly less than ideal meal. This is where we come in, trying to create the optimal meal for people.

What it does

The app serves to not only allow for better food recommendations for the diner, but also give restaurants a means of more powerful data analysis. From a very high level, the app serves as a means for massive data aggregation that aides both restaurants and users alike.

From the consumer side the app allows users to find a given restaurant and the associated menu. From here, the back end will also provide recommendations based upon a heuristic naive bayes algorithm.

How we built it

In terms of the tools we used: we built an app with a Flask framework frontend that is linked to a MySQL database and mobile development was done in swift. In order to aid our development, we leveraged a variety of APIs including locu (menus), yelp (restaurants), google search (autocompletion), and facebook-graph-api (future development and social network connectivity).

Challenges we ran into

In order to provide a reasonable proof of concept recommendation system that took into account the user's preferences we utilized the "yelp-academic-dataset." This led to some difficulties when dealing with the sparsely populated feature matrix and applying appropriate algorithms to achieve a reasonable result.

Accomplishments that we're proud of

For us, deciding on such a large project that included an extensive front and back end, in the end the experience was challenging yet really rewarding.

What we learned

For us some of the most valuable experience was learning to work as a team on such a large project (from both the front and back end). And this was truly our defining experience, where we're working with few hours of sleep on a large project, it honestly came down to delegation and team work that allowed us to finish.

What's next for chewyTime

For us, we still look forwards to continually improving and further developing the app and see where it takes us. This would mean hashing out more of the restaurant side of the application and building deeper functionality by leveraging networks.

Share this project:
×

Updates