Inspiration

Every day, citizens everywhere purchase groceries, commodities, and recreational items. Many of struggle to find the best prices and settle for sub-par prices. We believe that shopping should be a seamless experience with the ability to find the best prices at your fingertips. To conquer shopping, we created the pinnacle of shopping: a community fed shopping and pricing application. Its users will be the smartest and most connect shoppers anywhere. Reciebo is not just a tool for getting the best prices, but an asset that enhances every shopping experience you have.

What it does

Receibo takes user-input photos of receipts and parses the information to store in a database. Overall, this process takes three steps. First, the user uploads an image of their receipt which is fed into a Google Photo-to-Text API. Second, our algorithms pick out only the relevant information to store. Third, we push the information onto a database to further enable a proper info push to the frontend. As of 10/29, our capabilities include the name and address of the store, prices of individual items, and subtotals/totals. Moving into the future, we will add in a search function and improve and expand the algorithms for information identification. Additionally, we hope to implement Receibo to learn common market abbreviations and customer habits to reduce the user-check reliance and improve AI-UI connections.

How I built it

For the backend development, we used Python as our core coding structure. We integrated Google's Cloud Vision API to help with image to text processing. For frontend development, we used html to build our base website.

Challenges I ran into

The Google API was far more difficult to implement than we had previously anticipated. We expected to spend maybe 4 hours maximum debugging, but that quickly turned into 14 hours. Thankfully, we reached out for help and was able to make progress on integration.

We ran into frontend coding problems due to our limited experience with html and general frontend development.

Accomplishments that I'm proud of

We ended up getting a lot closer to fully integrating Google's API. While it was very stressful, I am proud of my team for holding out for the reward.

We finished most of the sorting and identification algorithms fairly early on. I was proud of my teammate for finishing it and immediately turning around to further assist with other areas of backend development.

We were able to get pretty far into html development, which was something I was not expecting. We overcame this by grinding through the problem until we got what we wanted.

We used custom designed logos and symbols for our website that were developed during the hackathon.

What I learned

I learned that Google's Cloud Vision API and really any API can be very annoying to integrate. However, integrating these tools is an essential part of technology and the work is well worth it.

I learned how to develop a full project in such a short time frame, including the communications and planning necessary.

Enough HTML/CSS to create a website.

Pirates of the Caribbean Theme Music makes the group twice as productive.

What's next for Receibo

In the future, we will work to further improve Receibo's algorithms and Google Cloud Vision API integration. Additionally, we will work to add profiles and a search mechanism to Receibo to further improve the user experience. We will also optimize both the desktop and mobile site.

Share this project:

Updates