PaperGist

Inspiration

To find an efficient method to convert articles into multimodal formats (for summary).

What it does

It takes in PDF files or webpage links as input, users will then be able to select what format they want the article to be converted into. This includes:

• Summary paragraph
• Bullet point form
• Keywords
• Image

How we built it

We incorporated OpenAi's GPT-3 API as well as google_image_search API into a PyQt5 GUI framework.

Challenges we ran into

Maintaining the functionality of the backend component of the UI was a challenge.

Accomplishments that we're proud of

• We are able to successfully generate concise summaries from a variety of articles in different format to better aid users in learning
• We successfully incorporated multiple APIs into a GUI container.

What we learned

• API generation and implementation.

What's next for PaperGist

• Conversationalistic AI requests i.e. giving more specific requests to the program such as asking it to generate powerpoint slides from a given article.

PaperGist was created during the 24 hour iNTUition v9.0 hackathon.