When technological advances reach educational systems, the result is oftentimes more student assessment. For example, new methods and research in natural language processing are often experienced by students in the form of automated essay scoring software. But instead of sorting and ranking students, what if we used the same or related technology to improve educational equity and opportunity? My research lies at the intersection of linguistics, artificial intelligence and machine learning, and is informed by critical perspectives on race and inequality from different disciplines. My current project uses AI to find textual patterns in college admissions essays written by Latinx students to reveal relationships between essay content and social context. As schools move away from standardized test scores in the admissions process, this work could help expose separate but related biases baked into the college admissions essay format. I hope this work leads to other AI projects to improve education.