We wanted to help users diet and stick to their budget by allowing users to take pictures of their meals to track dietary goals.
What it does
Our current implementation uses the google cloud Vision API to tag images of food, then searches the USDA nutrient database to find information on carbs, calories, fat, etc. for the food.
How I built it
We used many different tools, we started with python, bigquery, and cloud Dataprep. Then we decided to try a JS app with firebase and a rails app on app engine with the cloud vision gem.
Challenges I ran into
Our biggest challenge was in trying to use the vision API clients. We found that the vision api for firebase and the vision gem for ruby were difficult to understand from the documentation. The python library for the vision API was the easiest to use so we settled on that, but when trying to wrap our functions in Cloud Functions we ran into issues in setting up the routing tables for Cloud Functions endpoints.
Accomplishments that I'm proud of
We are proud we tried and learned so many different services and libraries on google cloud. We think this knowledge will make us much better hackers in future projects.
What I learned
We learned about google cloud SQL, App Engine, Dataprep, Bigquery, Cloud Functions, Vision API, and even the Natural Language API (to remove irrelevant terms from the object tags).
What's next for Nutrients and Stuff
We hope to use our knowledge of google cloud in future hackathons to better achieve our goals in the limited time as well as apply our knowledge to the workplace.