ReceiptIQ | Devpost

Inspiration

In many consumers' lives, the one thing everyone has in common is the need for food. This core component of making this food at home is groceries. You acquire these groceries by shopping. Every consumer needs to go grocery shopping. However, in many of these households, food is wasted because some groceries are forgotten or even go unused. That is the problem we decided to tackle.

What it does

Our application, ReceiptIQ, scans your grocery receipts to track your spending, monitor item expiration dates, and organize food by urgency. It then turns what you have left into personalized meal plans with recipes so you save money and also reduce waste.

How we built it

We built our program by combining OCR (optical character recognition) to extract item names, prices, and purchase details from the receipt. Our backend process identifies the items and the artificial intelligence API, then estimates the expiration dates based on the food type. Optical character recognition (OCR) reads the entire receipt and adds it to your total cost for budgeting purposes. Then the API uses the food list in the expiration list to suggest meals that could be created using the food items. Those food items have recipes that could also be accessed through a dropdown. In order to connect our backend with the frontend process, which is our webpage, using the Fast API.

Challenges we ran into

We ran into many challenges while we were coding. The main challenge we ran into was a bug when we tried to upload our receipt. The receipt OCR process was fine; however, the program was not navigating to the next page as intended. Then we realized we had put in a timeout counter, which was supposed to refresh the page if the upload process took too long. However, the threshold we put was super low. Therefore, we removed the timer, which fixed the issue.

Accomplishments that we're proud of

Accomplishments we are proud of include our ability to debug as we build our program. Another accomplishment is successfully using APIs to orchestrate the whole process of our application. Overall, we are proud of what we have accomplished in the time given, with minimal distractions and clear thought.

What we learned

We learned how to turn messy data into meaningful, useful information. Working with the receipt OCR showed us that extracting accurate information is not always straightforward. We also gained experience building from scratch, from the ground up. Beyond the technical side, we realized that working together and staying within our strengths are key to building a strong website.