We got the idea for foodie during lunch time. We couldn't figure out what we wanted to eat.
What it does
Foodie grabs the user generated photos from Yelp and displays them in a Tinder like format. Swipe left for "NOPE" swipe right for "LIKED" The app will keep a stack of foods you might like until you are ready to make your choice
How I built it
We built the app with Obj-C. We integrated a few API's to hep us verify the image as a food before we present them to the users. We are planning to also use the Magnet messaging app to create chat rooms so you and your friends can create the "LIKED" food stack together.
Challenges I ran into
Yelp doesn't have a built in API where we can pull the list of images from a given business. We had to parse it.
IBM's Watson is easiest to use with Node.JS, so we made a server for it. Also improving performance with Watson Facial Recognition was a challenge
What's next for Foodie
We plan to integrate our own database that will keep track of users likes. This will allow us to present images that are trending now. Yelp is great for displaying images that was great in history, but nothing that shows which food is trending. We felt that users might want ice scream if the weather is very hot.