Carbon Reciept

Inspiration

We wanted to know how the little things we do add up to affect the world

What it does

Scans a grocery store receipt and analyzes the amount of CO2 emissions released through that purchase. It also provides relevant suggestions to support decreasing our global impact.

How we built it

We used an OCR reading API to read the receipt and then parsed the items purchased into a Python script that uses a large dataset to calculate the carbon cost. This, through a Flask backend, is displayed on a React frontend along with some additional rating systems. We also used Moorcheh AI to create a chatbot that uses specific, credible research papers to help users improve their spending.

Challenges we ran into

We created our own model based on annotated receipts to detect the text; however, due to the lack of accuracy (despite being a great learning experience), we decided to use a public API service.

Accomplishments that we're proud of

Managing to connect many different services and our own algorithms to work together and create a cool product!

What we learned

Being able to use APIs and various tools such as Railway, Vercel, Moorcheh AI and so on. It really showed us what we can do with a huge toolkit of tools along with AI