The Impedance-Matched Multi-Axis Test (IMMAT) is a method developed to address the problems in conventional vibration test methods. However, in this method, some uncertainties must be addressed. For instance, the number of the shakers and their respective locations during a test need to be optimized to guarantee representative results. This optimization is usually done through a finite element analysis of the structure; however, there is not always a high fidelity finite element model available for the system. This usually happens when the structure has complex boundary conditions which are difficult to model or the material properties are unknown. The same is true for legacy equipment. If this is the case, data-driven models can be developed instead. To do so, the data available from the measurements on a number of locations on the structure are used to create a model. However, conventional data-driven models are limited to measured locations; thus, an infeasible amount of testing is required to expand these models. The recently developed Continuous Residue Interpolation (CRI) method yields a predictive data-driven model of the physical dynamical system, which addresses the limitations of the conventional data-driven methods. It turns out that the technique is powerful in both replicating the available measured data and predicting the non-measured data. It seems that CRI and IMMAT are natural fits for each other, and their combination can lead to an efficient procedure for dynamic testing and the overall general field of vibration testing. This research aims to inspect the use of CRI models instead of finite element models to optimize the IMMAT configuration. This can lead to the feasibility of carrying out an environmental test on complex structures where there are no analytical models available. Therefore, by testing the structure at a select number of points, a continuous model of the system can be developed, which can then be used to carry out a virtual test on a different combination of input and output locations to find the optimal configuration for a test.