Knowing who exactly is calling in a group of highly voluble (i.e., chatty) marmoset monkeys is not easy. That is where the recently published master thesis research of our now PhD Student Nikhil provides a novel solution. The machine learning pipline Nikhil and his coauthors developed is hierarchical, meaning the calls are first classified according to sex which narrows down the possible identities of the monkeys and only then as a second step the identities are assigned.