P. A. U. L.

Inspiration

The name Paul came to us randomly and we built around it.

What it does

P.A.U.L. aims to provide users with comprehensive insights on the effects of legislation on companies in the S&P 500 to help them save their time and ressources. To do this, P.A.U.L. analyses a bill, finds it's impacts on geographic and industry sectors, links these sectors to different companies in our database and mesures how large an impact the bill will have on them. P.A.U.L. then evaluates the impact on the companies thanks to an analysis of their clients and suppliers and proposes adjustments to the user's portfolio.

How we built it

We used various tools used across the industry, such as AWS, Bedrock, Python (with Flask), and HTML/CSS/Javascript, as well as an SQLite database.

Challenges we ran into

Token limit for our AI models SEC API would not let us import 10-Q and 8-K reports

Accomplishments that we're proud of

Finishing the project with a cohesive model

What we learned

We gained an in-depth understanding of AWS tools, as well as learning the specific needs of LLMs in an AI model. In a financial point of view, we learned what the different statements a company has to file with the SEC (10-K, 10-Q, 8-K, proxy statements), and how to read them.