Google Cloud x MLB All-Star Game

The brief.

Google Cloud wanted to show what its data science and machine learning stack could do for the fan experience, at one of baseball’s biggest live moments.

The work.

We built an AI pipeline that tracked the live game and modeled likely home run trajectories in real time, pulling from batting histories, weather data, and stadium topography. The model could predict where a home run would land down to the specific seat.

Vertex AI generated every line of content in seconds. An on-site team reviewed and lightly edited each message before it went to jumbotrons, scoreboards, and screens across the stadium.

Leading up to the game, the same engine powered mobile billboards driving around the host city. Headlines were generated on the fly from player stats, traffic patterns, and nearby landmarks.

My role.

Owned the production end to end: scope, budget, timeline, and delivery across strategy, engineering, and on-site operations. Vendor management, client management, delivered in 8 weeks.