Generative AI Innovations In Tennis: Kevin Farrar, Head Of Sport Partnerships For IBM UK, And Partnership Executive To Wimbledon/AELTC, In Conversation With Dinis Guarda

In an interview with Dinis Guarda, Kevin Farrar, Head of Sport Partnerships for IBM UK, and Partnership Executive to Wimbledon/AELTC, provides an in-depth look at how the generative AI solutions by IBM, like the latest ‘Catch Me Up‘ feature, for global sporting championships like Wimbledon, are reshaping the future of tennis with an enhanced fan engagement.

Generative AI Innovations In Tennis: Kevin Farrar, Head Of Sport Partnerships For IBM UK, And Partnership Executive To Wimbledon/AELTC, In Conversation With Dinis Guarda

The IBM Wimbledon partnership, which began in 1990, has been an era of new data and analytics driven sports. Wimbledon adopted the latest technology solutions to reach out to wider audiences with seamless and enhanced tennis experiences to its fans, and personalised playing styles and training methods for professional players.

Wimbledon gives us an opportunity to really showcase IBM facilities in a number of areas. Data is, obviously, at the heart of it, but there’s automation, security, AI, cloud computing, and on the consulting side, we work with Wimbledon to put a team that uses the IBM Garage Methods, agile design thinking etc. We are always looking for new data sources and from that data, both structured data and unstructured data,it has been a journey each year.”

Today, the tournament reaches a global audience of over 1.2 billion people across diverse geographic locations, encompassing 200 territories. Speaking about the IBM journey through three decades of Wimbledon, Kevin speaks with Dinis:

There are hundreds of millions of viewers around the world through the different channels who want people to have a feeling of being here without being here. In the app and on the website, there is a rich collection of visually blended images of what’s going on the ground between different players. These are some of the features. 

There’s been a lot of focus obviously over the last couple of years around generative AI. In 2017, we launched a feature where we were creating AI highlight reels of matches, listening to the noise of the crowd, looking at gestures of the players. We see a couple of stats that are exciting points in the match, in terms of Break Points or Championship Point, Match Point. We combine all those factors to give an excitement level to each of the rallies.”

AI enabled solutions from IBM: Enhancing fan engagement

In 2024, Wimbledon and IBM introduced a groundbreaking feature called ‘Catch Me Up’ that utilises generative AI to enhance the fan experience. This innovative feature provides personalised summaries of tennis matches, allowing fans to quickly catch up on the action they’ve missed. 

What we wanted to do was create an experience where fans, when they come into the app or the website and can read stories about their favourite players. Short form stories that give a bit of a preview ahead of matches and then also a post match summary of what’s going on during the match.

If you have a favourite player in the app or website, that’s the content that you will see. We also use geolocation to serve off these player cards and we also serve stories that are trending, that might be top seed, top rank players. So, we’re always trying to create fresh content so that when you’re going in we feed stories that you want to see”, Kevin told Dinis.

He also highlighted the powerful technology infrastructure of IBM watsonx that lies at the foundation of this feature:

What we do is we take an IBM foundation model, watsonx for example, with ‘Catch Me Up’ feature, we’re using a large language model and make sure that we’re using trusted data sources. We have a series of services that are running and then we also have a series of apps for example catch me up to keep our fans engaged.

The number of visitors to Wimbledon.com spikes massively in the leadup to and especially during The Championships. So, in order to scale effectively, we use a hybrid cloud infrastructure, a combination of IBM clouds and AWS clouds.

Commenting on IBM’s emphasis on the cybersecurity and governance aspect, Kevin tells Dinis: “It is a case of selecting the right model, selecting the right data training, and then there’s governance built in throughout that whole process.”

IBM: Shaping the future of AI in tennis and sports events like Wimbledon

As IBM continues to innovate, the focus is on making the Wimbledon experience more inclusive and diverse. Kevin highlights the role of generative AI in complementing the content created by Wimbledon’s team. “This is very much where we are complementing the content already created by the Wimbledon content team, enabling the generation of content at scale through watsonx and generative AI,” he explains.

Future developments include utilising multimodal models to process data from various inputs, including video and images. “The more data we can pick, the more insights we derive, and then we can create amazing experiences on top of those insights,” says Farrar.