As the tournament’s official digital innovation partner, IBM has helped the US Open attract and engage viewers for more than three decades. Year after year, IBM Consulting works with the United States Tennis Association (USTA) to transform massive amounts of data into meaningful insight for tennis fans.
This year, the USTA is using watsonx, IBM’s new AI and data platform for business. Bringing together traditional machine learning and generative AI with a family of enterprise-grade, IBM-trained foundation models, watsonx allows the USTA to deliver fan-pleasing, AI-driven features much more quickly. With watsonx, users can collaborate to specialize and deploy models for a wide variety of use cases, or build their own—making massive AI scalability possible.
Watsonx powers AI-generated tennis commentary
This year the US Open is using the generative AI capabilities of watsonx to deliver audio commentary and text captions on video highlight reels of every men’s and women’s singles match. Fans can hear play-by-play narration at the start and end of each reel, and for key points within. The AI commentary feature will be available through the US Open app and the US Open website.
The process to create the commentary began by populating a data store on watsonx.data, which connects and governs trusted data from disparate sources (such as player rankings going into the match, head-to-head records, match details and statistics).
Next, the teams trained a foundation model using watsonx.ai, a powerful studio for training, validating, tuning and deploying generative AI models for business. The US Open’s model was trained on the unique language of tennis, incorporating a wide variety of contextual description (such as adjectives like brilliant, dominant or impressive) based on lengths of rallies, number of aces, first-serve percentages, relative rankings and other key stats.
Beyond helping enterprise clients embed AI in their daily workflows, watsonx helps them manage the entire AI lifecycle. That’s why the US Open will also use watsonx.governance to direct, manage and monitor its AI activities. It will help them operationalize and automate governance of their models to ensure responsible, transparent and explainable AI workflows, identify and mitigate bias and drift, capture and document model metadata and foster a collaborative environment.
Using watsonx to provide wide-ranging Match Insights
The US Open also relies on watsonx to provide Match Insights, an engaging variety of tennis statistics and predictions delivered through the US Open app and website.
For example, the IBM Power Index is a measure of momentum that melds performance and punditry. Structured historical data about every player is combined with an analysis of unstructured data (language and sentiment derived from millions of news articles about athletes in the tournament), using watsonx.data and watsonx.ai. As play progresses, a further 2.7 million data points are captured, drawn from every shot of every match. This creates a rich, up-to-the-minute data set on which to run predictive AI, project winners and identify keys to match success. The Power Index provides nuanced and timely predictions that spark lively engagement and debate.
When a tournament draw is released, pundits and fans often assess each player’s luck and path through the field: do they have a “good draw” or a “bad draw”? This year, IBM AI Draw Analysis helps them make more data-informed predictions by providing a statistical factor (a draw ranking) for each player in the men’s and women’s singles events. The analysis, derived from structured and unstructured data using watsonx, determines the level of advantage or disadvantage for each player and is updated throughout the day as the tournament progresses and players are eliminated. Every player has their draw ranked from 1 (most favorable) to 128 (most difficult). Fans can also click on individual matches to see a projected difficulty for that round.
Based on the AI Draw Analysis, users of the US Open app can explore a player’s road to the final and the difficulty of a player’s draw. The AI Draw Analysis feature shows potential matchups, informed by player data derived from the Power Index. For matches in progress, fans can also follow the live scores and stats provided by IBM SlamTracker.
A new era of scalable enterprise AI
Through their longstanding partnership, the IBM and the USTA collaborate to explore new ways to use automation and AI to deliver compelling fan experiences at the US Open. This year’s innovations demonstrate how watsonx can help organizations quickly and effectively implement both predictive and generative AI technologies. Through a collaborative, centrally governed environment that empowers non-technical users to make the most of their organization’s high-quality data and leverage foundation models trained on IBM-curated datasets, watsonx opens the door to true AI scalability for the enterprise.