MedScale

Inspiration

We got inspired through the various science courses we have taken in our university careers! Many of these courses require us to research various articles and papers and analyze and review any questionnaires and surveys present within them. However, there is no easy way to do this. Unfortunately, this research becomes a very time consuming task to do as we have to read through long articles, trying to find these questionnaires/surveys.

What it does

Our tool provides a temporary solution to our problem. It will read any PDF document and will give you all the instances where there are questionnaires/surveys.

Our website provides a vision for our long term solution. Here, the user will be able to input any topic they want, and our site will be able to generate filtered results, presenting the questionnaires/surveys to them. This is an amazing solution as this will not only allow the user to view the questionnairs/surveys, but they can compare them with other ones present.

How we built it

Our tool was built through Python on Conda. Our tool will accept a PDF, convert it to a text-file, then search the text-file for key words of 'questionnaire' or 'survey' and will display these instances along with the line and line number. One thing we are working on for our site is to provide a summary using cohere. Unfortunately, the prompt that cohere requires for the code that we were trying to use, will only accept a string that is not very long.

Our website was built using Figma. We were able to create different desktops and simulate the search process, account process and about us process of the site. We also did create something that is very user friendly, unlike our tool, and something that is less intimidating so that anyone, even those with no coding background can use.

Challenges we ran into

Some challenges we ran into was the summary part of our code. Cohere has some code on their site that we can use, which will summarize a prompt for you. However, the prompt has constraints; it needs to be a string and it cannot be very long. This was difficult for us as medical articles/journals are quite lengthy.

With Figma, we were learning as we were going. We did face some problems with creating a good UI while trying to incorporate everything the site has to offer. Some other challenges we ran into was the fact that Figma is mainly a UI platform. So it is hard to test out UX with it due to the lack of user interaction.

Generally, we are not very proficient in coding. So during this hackathon, it was a bit of a learning process as we were learning code and UI/UX on the go. This did prove to be a fun process though.

Accomplishments that we're proud of

We are so proud that we actually produced an idea that works, and provided both a UI site and some code to show that it can be done in the back end. We are also proud of the fact that we were able to create both short term and long term solutions to our problem. The short term solution could even be used today!

What we learned

This whole hackathon was a learning process as we both are not big coders. However, we were able to (re)learn python and create a very usable tool that does cut down our research time in half! We were also able to learn more about UI/UX and thoroughly explore that area through watching videos online and doing research.

What's next for MedScale

We would definitely love to see our site actually work one day! The next steps would be to invest some time in recruiting people that could help us more with the back end processes of this.

Built With

• cohere
• figma
• python