59% of Americans look at the ingredients and nutrition label, but not everyone has the time or knowledge to cross-reference all the ingredients. We take all the heavy lifting out of the way by having a natural language processing algorithm to quickly identify ingredients and risks/benefits.
What it does
Take an image of an item's nutrition facts on our mobile app -> Our back end servers use GCP to identify text -> If a hit is created in our database -> Tell the user of nutrition/ingredients that may be positive or negative
How I built it
Our vision and natural language processing APIs came from Google Cloud. Our app front end was made with React-Native, while our back end was made with Flask.
Challenges I ran into
We didn't have extensive React knowledge, so it took us a while to make the app exactly like we wanted. Also there weren't many good databases for ingredients and their side effects so we used Web scraping and manual labor to make our own database.
Accomplishments that I'm proud of
Making a React-Native App.
What I learned
Flask React-Native GCP
What's next for SnapStats
- On the fly/immediate translation
- Bigger database
- Personal Blacklist
- Ability to crop out ingredients
- Potential to analyze items without ingredient labels
- Fine tune the data (ratings)