The idea is to implement a more base-level segregation of data inside the database to make it less likely that a programming error could lead to data leakage (e.g. Team A seeing data of Team B).