<p>The injection moulding process is a widely used method for manufacturing plastic parts due to its high productivity and ability to produce large quantities of quality components. While injection moulds are typically made from P20 steel alloys, polymer moulds (soft tooling), particularly those produced through additive manufacturing, offer an interesting alternative for small production runs. However, the lower mechanical strength and thermal conductivity (typically 200 times lower than P20 steel) of polymer moulds make defining injection parameters a more complex task, requiring experienced operators to prevent tool failure during trials and injection. This article presents a methodology for obtaining ideal injection parameters using statistical and optimization methods. A full factorial design was used for numerical simulation setup, analysis of variance to assess parameter influence on warpage, response surface graphs to map the solution space, numerical simulations in Moldflow, and statistical tests in Minitab. Two optimization methods, deterministic (gradient-based) and heuristic, were compared for effectiveness (result convergence and computational time). The gradient-based method demonstrated better results and faster solution times, while the heuristic method converged to a local minimum. The optimization technique resulted in a 1.87% reduction in part warpage compared to the best factorial design result. This study provides insights for operators on configuring injection parameters and improving the dimensional quality of plastic parts by reducing warpage.</p>