PREMSCOUT

Inspiration

We as a group are known for our interest and involvement in the sports world so when HackUTD and Pinata presented us the opportunity to unleash our creativity, PREMSCOUT was a natural thought. Creating a tool that helps us not only learn a lot about software development and computer science but also aids us in our entertainment in Premier League Fantasy was a privilege.

What it does

PREMSCOUT consolidates loads of Premier League data about players and delivers AI powered calculations to help Premier League Fantasy players make smarter decisions. It also boasts a slew of tools to help visualize and learn about players in different perspectives.

How we built it

First, we came up with a design and starting refining it through Figma, making sure our visions were aligned. For the Frontend, we utilized React and Tailwind CSS. Pinata was the backbone of our data-centric project as it served as our extremely easy to use database. For our AI Model, we used scikit-learn, numpy, pandas, sklearn, and joblib. Finally, for data scraping from the web, we utilized Python, Selenium, and ChromeDriver.

Challenges we ran into

Due to our inexperience with project based development and the traditional workflows projects like this entail, the learning curve we had to experience for every step of development was frustrating but extremely exciting to conquer. Concepts like AI models and data scraping set us back in time but with the support of each other made the result much more gratifying.

Accomplishments that we're proud of

Making an attractive website within the given time limit that achieves everything we set out to do. Not only do we have AI-calculated projections, but we've created an intuitive tool to learn about players in a variety of ways.

What we learned

We learned what it takes to work in a team, be responsible for features, and how to work past our limitations.

What's next for PREMSCOUT

Since this was the first time, we've created a project like this, we missed out on a lot of opportunities to keep our code more modular and future proof. Not only that, but the lack of automation on several parts of our design definitely leaves room for further development and reconfiguration to get closer and closer to the tool we dreamed of when we started this hackathon.