Happy Tummies

Logo
HomeScreen
HomeScreen2
Query

Inspiration

My gym friend alwayskept complaining about how hard it is for him to bulk since his mom doesn't allow him to buy meals from outside

What it does

So we thought, why not we create something that solves that problem which lead us to build Happy Tummies. It takes a peek at your fridge and then gives you tasty and healthy meals based on what you have

How we built it

We let user upload a picture which gets stored in AWS S3 and then gets proccessed by Clarifai API which tells the food items and confidence level. We then filter it out and make a list of items that is most likely to be in the fridge. This list then gets fed into Azure OpenAi completion API which returns a text of the recipe. We then parse the text and arrange it in an organized manner and display it back on our website.

Challenges we ran into

Finding an API that can detect food and getting around the costs of OpenAI with no options of getting a chat API. Getting a downloadable link in S3. Also our statistics exam.

Accomplishments that we're proud of

A working beta model despite the time constraints and exam pressure. Coming up with something original and something that actually solves a problem

What we learned

APIs. Tons of APIs (especially AWS :P) Budget restrictions, finding work arounds. Guiding through documentation. Asynchronous calls and accomodating size restrictions.

What's next for Happy Tummies

We are planning to build a virtual fridge and a food calculator which will tell the users what essential items they can buy. We are also planning to connect them with local vendors and give suggestions on how they can dispose their waste responsibly