Case Study: Kabam improves game quality and ML debugging with Weights & Biases

Kabam Keeps Game Development Front and Center with Weights & Biases

Kabam, a leader in immersive mobile games, sought to ensure a high-quality, bug-free player experience. Their challenge was shortening feedback loops and quickly identifying discrepancies within their complex machine learning pipelines used for game development. They partnered with Weights & Biases to tackle this.

By implementing Weights & Biases, Kabam's team used the wandb.log function to visualize and debug their ML workflows. This provided dashboards that allowed them to quickly solve problems and efficiently compare experiments. The solution helped them build a reliable AI game design assistant, which in turn enabled their game team to test new features and catch disruptive issues early, ensuring a seamless experience for players.

Kabam

Eric Chou

Senior Software Engineer