Most zone or room level control-oriented models are developed with opaque and varying workflows. The workflows vary depending on whether the models are white or black box. In this paper the focus is applied to inverse modelling. The contribution of this paper is the application of a simple prescribed workflow for generating inverse models of a group of offices that can be applied by those less experienced with inverse modelling. The goal of the models is to predict indoor air temperature at the next time step for each room with reasonable accuracy. Measurements were taken in 27 offices for a period of one month. These data were used to fit and validate office level control-oriented inverse models using the MATLAB System Identification Toolbox. The maximum mean absolute error (MAE) for any office over the two-week validation period was 0.31°C and the average MAE for all 27 offices was 0.15°C.