This project is our submission to these sponsors' challenges: Goldman Sachs, UTD Student Government.
Inspiration
Financial hardship is something everyone wants to avoid at all cost, especially college students. Thus, we would love to bring the functionalities that Comet Trade offers to others that they are able to reduce as much financial burden as possible.
What it does
A student-friendly investment platform that aims to maximize profits and minimize risks in trading by using AI algorithms to make investment predictions and strategies. The goal of the application is to promote fiscal responsibility and financial intuitive in college students as well as midgate risk for financial institutions using AI/ML and big data.
How we built it
To implement the web app, we utilized ReactJS for Frontend development. We used MongoDB for the database and Flask for the API and middleware. We used yfinance API to retrieve, restore stock information as well as current data and news for securities. We used tf-idf to weight and cluster terms based on relevancy and used BERT to analyze sentiment of stock data from twitter and news.
Challenges we ran into
Lack of preperation, lack of time, weak communication, backend and frontend integration, different looks of UI in different dimension.
Accomplishments that we're proud of
We are proud that we made a very good looking UI and applied advanced algorithms to elevate the app even though we are in short of time.
What we learned
We learned about stocks and finance API. Additionally, we also gained more knowledge on advanced AI algoritms outside of class.
What's next for Comet Trade
In the future, we would like to implement the time-series stock forecasting model for our app as well as conducting volatility predictions for our data. Moreover, we also would want to build more functionalites, transactions, for example.
Log in or sign up for Devpost to join the conversation.