Inspiration

It's always hard to find a place when you're going out with friends. Especially when everyone likes different foods. That's where Arcuisine comes in. We use machine learning to tackle the difficult problem of finding the best possible place for everyone based on your individual likes, ratings, and past experiences with other friends.

What it does

Arcuisine finds local restaurants for your group of friends to go to based on all of your likes. You create a profile and specify what cuisine types you personally like or dislike. It then allows you to connect with friends in a group, and once all of your friends are partied up, it will use your profile information to suggest some of the optimal results in the area for where you should eat!

How we built it

We built Arcuisine in three main components. Firstly, there is a server backend that uses node.js and MongoDB that provide for user information storage to the other components. Next, there's a web-based UI written using Jade which provides a more detailed view of each user's profile and the groups that they are a part of. The final component is an iOS app created using Swift which allows mobile users to connect in the same ways as the desktop users. Additionally, this provides an environment for better localization in the future.

Challenges we ran into

Some large challenges included learning Swift and building our first app without reliable access to compatible hardware.

We also found initial resistance interfacing between our server and the device as well as setting up and managing the database.

Accomplishments that we're proud of

The communication works quite reliability between all three of our components when we only had a short time period to configure it.

What we learned

We learned quite a bit from this project ranging from familiarization with database management, Swift/XCode, and numerous helpful libraries. This was both of our first experience building an iOS app.

What's next for Arcuisine

We want to add machine learning techniques to produce the optimal list of restaurants and maps to show these places, which we weren't able to do due to time constraints. With more time, we would add the ability to actually like/dislike cuisine types and rate past restaurant visits.

Built With

Share this project:

Updates