Non-parametric processes and graphical models are two tools for working with high-dimensional data. Non-parametric processes allow one to avoid restrictive assumptions about the shape of a distribution. This is especially useful in large dimensions because there is often insufficient quantities of data to validate assumptions. Graphical models ease computational burden via "conditional independence relationships", which express complicated interactions as a series of smaller-order relationships. This talk is part of a growing literature that seeks to combine the benefits of non-parametric and graphical approaches. We focus on developing a graphical version of the well-known Dirichlet Process, including construction and extensions to current applications to accommodate graphical models.