SummarizeMate

Inspiration

We found reading long handouts annoying and a wasteful use of our time. At times we would get handouts that had no digital copies which would make things a lot more difficult.

What it does

The program converts your handout/notes to text using advanced OCR. This text is processed to make a summary and given to the user.

How we built it

We leveraged google cloud vision and cohere APIs for OCR and summarization. We use taipy to make a frontend that recieves inputs that are put into the API calls and then

Who did what

APIs, Backend and MongoDB - Om and Rehman Taipy, frontend and Autho - Abhishek and Aakash

Challenges we ran into

Leveraging advanced APIs like Cloud Vision and cohere was something new. It took us some time to get into the flow of getting the API key, using it correctly, initialize the client, etc. TaiPy was a python library we had no knowledge of prior to this hackathon and it was also a module that didn't have much resources on the web. Figuring out how to use it to code our web app (frontend) and integrating it with our model of the image to text summarizer was some of major the challenges we ran it to. MongoDB and AuthO was also something completely new to us and we underestimated the time it would take to implement the technologies.