For a known player in Belgium, we made a full web platform to extract, visualize & aggregate soccer player data using various new techniques in the Artificial Intelligence field.

The developed website, with a micro services backend, was the capstone project for our bachelor thesis. It was a whole year during group project with all seniors collaborating in scrum squads, in agile sprints (Not full time, we had 5 other courses this year). The goal of the project was to create a platform where soccer analysts, trainers and coaches could get automated insights & predictions in their competing behavior.

The result was a platform where you could ingest video data, health statistic & competition results. From the video data, positions & key-moment replays we’re automatically extracted, or annotated. Humans could also take part and manually intervene. Other modules were used to make predictions about a players future health (e.g. injuries), chance to score or any other combination of the data in the system (A user could define their own prediction models). Tactics where visualized in both 2D & 3D, think Fifa. A chatbot was added to aid the enrolling process and improve the usability of the platform. The platform also included functionality to plot various graphs.

While the project was sometimes chaotic and stressful (I’m still convinced some issues we’re accidentally created by our teaching assistants to see how we would cooperate), it was a very meaningful experience and trajectory to make. Especially since it was a real project!

Together with 3 of my fellow students, I had the honour to give the final demo presentation.

ps: We needed to sign a Non-Disclosure Agreement (NDA), so I can’t elaborate more then this on the system.