Philip originally developed the foundation of the AI element of the project for his undergraduate dissertation, focusing on automated ball tracking. The next year, Jeremy and Liam further developed the idea building the hardware and companion app for their master’s theses, while Phil was at Cambridge studying for a masters in AI and Machine Learning. Now that all three have graduated, they’re back together working full time on Pitchwise.
The team have built and tested working prototypes with several local teams across Bristol and are now starting a much wider pilot programme. Alongside this, they have been refining their AI models for player tracking and collecting the crucial data needed to further advance them.
The Isambard-AI access (5,500 graphic processing unit hours, valued at £25,000) through the UKRI AI Research Resource [AIRR] programme supported by Innovate UK, will allow the Pitchwise team to train and scale their computer-vision models far beyond what’s possible on standard cloud hardware — enabling faster iteration, deeper analysis, and more advanced player recognition.