Although apparently simple, the polycondensation reaction leading to polyazomethine is difficult to control because of its equilibrium character, the conversion degree being influenced by a series of parameters. The reaction between a siloxanediamine, 1,3bis(3-aminopropyl)tetramethyldisiloxane, and terephthalaldehyde was performed here in solution (in tetrahydrofuran) without byproducts removal and in absence of any catalyst or pH modifier. Different conditions (co-monomers ratio, dilution, and temperature), considered as input parameters for the process modeling, were varied according to a pre-established experimental program. The viscosity of the reaction mixture was chosen as output parameter, being monitored with a Haake Viscotester 7 Plus-L. The process modeling was performed using a hybrid combination of artificial neural networks and differential evolution algorithm, the last one having the role of developing the neural model in an optimal form. The simulation results showed that the methodology provides accurate results, the model predictions being in close correlation with the experimental data.