Teamwork is critical to engineering professional work. While some aspects of teaming with engineering students are well understood and implemented into instructional tools, tools for handling student teams dealing with implicit and explicit racism, sexism, and homophobia are infrequent. Instructors of large undergraduate courses need tools to help make team-level marginalization visible at the classroom level to interrupt discriminatory or marginalizing behavior amongst teammates, and to model allyship so teammates learn how to interrupt others' marginalizing behavior when instructors are not around. This paper describes the broader project, and describes some early results, focused on an algorithm that can help identify teams engaging in marginalizing behaviors against minoritized students, whether minoritized by race, gender, nationality, LGBTQ identity, or other categorization schemes. We describe how the algorithm is proving useful to identify student teams to focus on for analysis to answer some of our research questions focused on how engineering undergraduate teams marginalize minoritized members, and illustrate one such analysis. We describe our continuing work on the broader project.