how it works

  • The latest iteration of doofieBot lurks on Slack, its backend built with Flask & powered by Amazon EC2.
  • doofieBot makes use of IBM's powerful Watson Visual Recognition API to interpret pictures of delicious food queried by the user.
  • After intelligently recognizing the food, doofieBot calls on Yelp Fusion to generate a list of nearby restaurants or businesses that serve or sell what you are looking for!

/doof

how it's built

how it was done

  • doofieBot was built by a collaborative team of students from UBC (University of British Columbia) and UTSC (University of Toronto in Scarborough) for Hack the North 2017 (15-17 September).

more background

doofieBot was inspired by the fear of missing out on good food. what happens if your friend snapchats you a picture of a particularly good meal, and you start craving it too? by intelligently identifying the food and harnessing the power of a good restaurant analytics API, /doof is here to save the day!

challenges & accomplishments

The biggest challenge we faced (and are still facing) is the integration of doofieBot into the Facebook Messenger Platform, slowed down by the daunting task of implementing an SSL server against budget & time constraints. Looking forward, we are still looking to make better use of Google Firebase in order to properly implement the integration. Aside from that piece of unfinished work and some more considerations we have in the future for a better web UI, this was a great project overall & we are very proud of the pipeline that we have designed and executed in the span of 30 hours.

Share this project:
×

Updates