Julia Dressel is a software engineer on the Lantern and Public Data teams at Recidiviz. She joined Recidiviz as their first full-time employee after working at Apple as a MacOS developer. She is the co-author of the Science Advances study “The accuracy, fairness, and limits of predicting recidivism”, which received global media attention for exposing that a popular algorithmic risk assessment tool does not outperform human prediction. When she’s not thinking about how systemic biases are reinforced through technology, she’s either skiing in Tahoe or playing flag football in the SF Womxn’s Flag Football League. She holds a B.A. in Women’s, Gender, & Sexuality Studies and Computer Science from Dartmouth College.