For the last 8 months I've been developing a site I call monkeemath.com. The basic premise of the site is I scrape stock market related subreddits looking for comments that include stock market tickers, e.g APPL. The comment then gets sent to ChatGPT's API endpoint where I ask it to determine whether the comment is bullish or bearish on the outlook of the ticker it mentions.

You can do a lot of different stuff with this data, the obvious one is mapping the mention volume/ sentiment over the price movement of the associated stock. You can see this done here -> www.monkeemath.com/pages/movementgraph

Some other features the site includes: - Raw comment viewing: View the comments the site scrapes in their raw form, filter by ticker, subreddit, sentiment etc - User predictions: Users can make accounts and make predictions on the price movement of a given stock, you can then view other users' predictions, their explanations and how accurate their predictions have been in the past. There’s also a leaderboard on the homepage showing the most accurate users. - Article Sentiment: Scraping articles from over 500 different sites and using LLM sentiment analysis to determine mention sentiment (click on a symbol to view articles that mention it) - MonkeeMath API: Programmatically access all my data (for algo traders or other interested parties