ZapFA

Inspiration

As college students, we have experienced the need for financial independence and the importance of investing in stocks. Hence, we wanted to make a virtual assistant, which would empower individual traders.

What it does

It takes a ticker symbol the user wants to invest in. After getting that, we fetch data with respect to that ticker symbol using the official yahoo finance API and pass it to a machine learning algorithm. The algorithm aims to predict the closing price values for the next 30 days using linear regression. As the machine is predicting for each day, we optimize the prediction in real life in accordance with the previous data to make the best predictions. It returns the data back to the website and displays a graph, company information, and all kinds of useful data.

How we built it

We built it using programming languages like python which enabled us to build an API around it. After that, we used the flask module to set up our website and display the information we got from using Yahoo Finance. Moreover, we used canvas.js to make the graphs. And we used Heroku to deploy the website.

Challenges we ran into

One of the main difficulties was how we were going to connect the backend to the frontend. Other than that, the graph plotting was a huge mess as we didn't know which JS module we should use. Hence, it all got resolved by the trial-and-error method. In between programming our files, there were a lot of hidden difficulties we ran into as well.

Accomplishments that we're proud of

We are proud of the hard work that we have put into this product. And since this is our first Hackathon, we couldn't be more proud of our product. Also, we collaborated a lot over these two days and we are really proud of how everything turned out.

What we learned

Personally, I learned that every minute of planning saves an hour of debugging. Additionally, I think teamwork and having good spirits help keep the team moving on throughout the night. Lastly, I would say that a solid division of labor keeps the team running at maximum efficiency. However, when multiple people work on the same thing, conflicts may happen which slows down progress.

What's next for ZapFA

Next, we really want to focus on refining and expanding our stock prediction algorithm. Currently, we use linear regression - however, there are a wide variety of stock analysis algorithms to pick, choose, and implement. Imagine the end-user having that power at the end of their fingertips. When we have enough time and resources, our next step is to let the user create their personal trading bots and test their strategies, creating a community of young and aspiring investors.