Geometrical inaccuracy from shrinkage and residual stress-induced deformations are key sources of defects in Additive Manufacturing (AM).  In most AM processes the CAD model is represented by an STL file which is sliced into layers. The Common Layer Interface (CLI) and StereoLithography Interface (SLI) files are two layer file formats which store this slice information and can be data mined to analyze sources of error.  By comparing the original STL file against a ‘stacked’ model based upon SLI and CLI files, a “theoretical" part accuracy can be created.  In addition, these SLI/CLI files can be combined with detailed process parameter knowledge such as scan pattern, laser power, scan speed, etc., to enable a numerical prediction of part accuracy and deformations on a layer-by-layer basis.  This paper introduces a research project which is developing software and hardware tools to enable prediction and measurement of part dimensions on a layer-by-layer basis with the goal of real-time part quality validation and closed loop control.