Motivation!

In the past I have experimented with building stock bots which able to buy and sell stocks however these algorithms were powered by statistical analysis and were not user friendly. Today I wanted to do the opposite...

How it works!

Powered by Gemini, this stock bot works by reading recently published articles about stocks and crypto currencies and making a list of the best assets according experts. Then the bot fetches live data to verify claims that the given asset is a good investment. If the current price of the investment is on the rise then the bot will purchase the investment without the need for human intervention.

User friendly GUI!

The extremely user friendly GUI for the bot allows the user to browse what articles the bot read and what assets it decided to invest in as a result of the research it conducted autonomously. It also allows the user to set limits on the amount of capital the bot has access to and manually start a buy event (to show the bot actually researching and buying an asset).

Challenges faced.

The most difficult challenge that I faced was converting Gemini's result what to invest in into a function call that can gather more data on that asset and to ultimately buy that asset. While I was not able to fully solve this issue I was able to significantly increase the reliability by parsing through the result converting the string into a readable json file.

How it was built.

The project was first build by starting with getting Gemini to find news articles. I then worked on developing functions to buy and sell assets using alpaca API. Next, I worked on modifying the code which found news articles to find recent news articles on assets [stock or crypto] and to create a list of the best assets to invest in. Finally I developed the GUI which to visualize the bots actions and refined the bots actions to research verify and invest in assets autonomously.

Built With

Share this project:

Updates