The way we store our data sometimes makes a lot of difference. At Trax, we store our data in relational tabular form, and while this data representation is useful for certain queries and applications, we had a hard time calculating several KPI's and answering certain business questions. In this session, I will present how using a complementary graph representation of our data helped us utilize the geometrical aspects of the shelf to solve problems in a much easier and efficient way and opened a door to new capabilities.