One of the growing areas of interest in the context of biological and medical models is the question of how to personalize a model to a target organism or patient in order to improve model predictions. This question is inherently tied to data assimilation (DA): the process of calculating the “optimal” combination of observational and model data which determines the next model state while taking into account the errors present in both sources. Most existing DA methods have evolved from the context of numerical weather prediction and are tailored toward ordinary or partial differential equations models. However, little work has been done to extend the application of this methodology into the context of more complex models, including stochastic and multiscale models. In our work, we consider how to extend a well-known DA method, the ensemble Kalman filter, so that it may be applied to some stochastic models of increasing complexity. We explore the benefits and limitations of this method in doing so and discuss the open questions that still exist within this area of research.