As university students, we often find that we have groceries in the fridge but we end up eating out and the groceries end up going bad.
What It Does
After you buy groceries from supermarkets, you can use our app to take a picture of your receipt. Our app will parse through the items in the receipts and add the items into the database representing your fridge. Using the items you have in your fridge, our app will be able to recommend recipes for dishes for you to make.
How We Built It
On the back-end, we have a Flask server that receives the image from the front-end through ngrok and then sends the image of the receipt to Google Cloud Vision to get the text extracted. We then post-process the data we receive to filter out any unwanted noise in the data.
On the front-end, our app is built using react-native, using axios to query from the recipe API, and then stores data into Firebase.
Challenges We Ran Into
Some of the challenges we ran into included deploying our Flask to Google App Engine, and styling in react. We found that it was not possible to write into Google App Engine storage, instead we had to write into Firestore and have that interact with Google App Engine.
On the frontend, we had trouble designing the UI to be responsive across platforms, especially since we were relatively inexperienced with React Native development. We also had trouble finding a recipe API that suited our needs and had sufficient documentation.