"After our team's intervention, the model's accuracy improved across all demographic groups with less than 1% variation between them, and the company adopted our testing protocol as the new standard.

"I was assigned to audit the company's new AI recruiting tool to ensure it was free from gender bias before deploying it to our clients."

"I implemented a new testing framework that incorporated diverse data samples representing different demographics, and then recalibrated the model's parameters to reduce bias."