Wallstreetbets Analyst

Main Website Screen
Main Website Scrolled Down
WSB vs S&P 500 Portfolio
Chatbot 1
Chatbot 2

Inspiration

Initial Idea Our initial idea had to do with analyzing posts from popular politicians to see if their sentiment followed the actual stock value, and how much money it would have made Updated Idea Due to Twitter API costs, we switched to a well-known Reddit community, WallStreetBets, which is known for risky, high-reward stock picks, aggressive trading strategies, and a rebellious, meme-driven culture that challenges traditional Wall Street norms.

What it does

Our program first collected all the posts from the past month on the WSB (WallStreetBets) Reddit community, with information such as the title, body text, upvotes, date posted, and top 10 comments to analyze if the post's sentiment was supported by the community. Then we used Gemini API to extract the stock ticker and post sentiment to see which stocks were to be bought and which to be shorted. Next, we ran a simulation through our data to simulate a portfolio for the WSB community in the pasts month. We also created a website to hold this information, with the WSB portfolio being compared to the S&P 500. This website also hosts a chatbot that can give stock recommendations based on our analysis of each post's sentiment.

How we built it

We used Gemini to analyze post sentiment and for the chatbot. We used flask for the backend of the website and react for the frontend. The simulation was done using Python.

Challenges we ran into

We ran into problems with different ways to parse the tickers and sentiment from the post. We tried using regex and nlp but decided to use Gemini because those methods were not very good. The front-end implementation and simulation implementation were very difficult as well.

Accomplishments that we're proud of

We are proud of utilizing the Gemini API for post-analysis and the chatbot. We are also proud of being able to complete this project in 24 hours and have a functional backend and frontend as well.