StockBot

stock buying
stock data

Who are we?

Hi, today our project focuses on buying and selling stocks based on future predictions. With a machine learning model imported via Kaggle, we can decide when to buy, sell, or hold stocks.

Inspiration

To make trading easier to do without the hassle of thinking too hard about it.

What it does

It'll allow the user to trade stocks of their choice automatically for them. Given a stock to monitor and a few parameters to act as threshold (like a limit to how much money we're willing to lose), the bot will run a prediction model to see the prices of stocks for the next few days. Based on these predictions, the bot will decide to hold, buy or sell the stock. The bot can also be set to do this on a daily interval (i.e. it'll check predictions and trade stocks twice a day if asked).

How we built it

We've built this project through a multitude of code sections. First off, we pull data for predictions using an API named Tiingo. Then, the data is sent into the AI Model imported through a series of file readers. The Kaggle Model that will be used to reformat the data and output closing dates and prices. Then, we could make a request to a market simulator (the actual market is closed on weekends) to buy and sell stocks

Challenges we ran into

The authentication token changed in real time for each page as we adjusted each stock, so scraping the token for it resulting in some issues with trying to make a proper request to intercept the network flow.
Keep the application running by applying different methods since once the code was finished.