ReSearch

Logo
Search For Papers
List of Papers
View Research Paper PDF
Open AI Explanation on prompt
Logo

Inspiration

People publish a lot of fascinating research out to the world, yet the tools to consume this research are quite primitive. It is also really hard to understand the research papers.

If you're coming from a research background, you must have stumbled upon Google Scholar's messy UI, scrolled on websites blocked by paywalls, or wanted to download the research paper.

What it does

Our platform is the go-to platform for accessing and understanding research papers. We provide a fascinating interface for users to search for papers and return a list of papers with their abstracts and a direct pdf link in a prettified format.

Since research papers are generally difficult to understand, we have added a explain paper feature using Open AI's GPT-3 model. Users can simply copy-paste an excerpt and our primed model will translate it into an easy and understandable form.

How we built it

Next.js, Tailwind CSS and Framer Motion for Frontend
Used Libraries like React Hot Toast and React Rough Notation to make it look interactive
Added authentication within the app using Auth0
Flask backend fetching papers from Cornell University's extensive Arxiv API
Tuned the GPT-3 Model in accordance to the context with our training text

Challenges we ran into

We faced quite a lot of challenges in building the hack over the weekend.

Integrating Flask with a Next.js client running with TypeScript type support.
Adding Auth0's secure authentication into the app.
Embedding PDF and passing to the GPT-3 Model

Accomplishments that we're proud of

We are proud to have implemented the idea we thought of at the beginning of the hackathon.

What we learned

Learnt so many new tools over the weekend like Next.js integration w/ Flask and Open AI
Worked on Auth0 bugs as well
Built a simple yet elegant UI using Next.js
Embedding PDFs into the app