The structural modifications of a commercially available poly(lactid acid) grade were induced through reactive extrusion using a multi-epoxide reactive agent in a pilot plant. The statistical nature of the chemical reactions led to the generation of several types of non-uniform molecular architectures. Even though conventional spectroscopic (NMR) or chromatographic (SEC-static light scattering) techniques are placed at the forefront of the molecular characterization, both methods usually failed in characterizing non-uniform structures. In this study, a promising approach was applied to automatically classify modified PLA samples. It is based on the analysis of FTIR spectral data by means of multivariable methods, including feature extraction and classification algorithms. The fast and accurate results presented in this paper show the potential of the proposed approach.