The study aims at investigating the design and performance improvement of 3D printed structures using Fused Deposition Modelling (FDM) and the PLA+, ABS and carbon-fibre-reinforced nylon (PA-CF) materials. The paper analyses the influence of some of the major parameters of the process to the tensile strength, surface roughness, wear rate and dimensional accuracy: layer thickness, raster angle, and infill density. Experimental tests were performed in the ASTM specifications and the observations were determined using Response Surface Methodology (RSM) and hybrid Artificial Intelligence (AI) optimization methods such as GA-ANN, GA-ANFIS, GA-Fuzzy model. PA-CF had the greatest tensile strength (62.8 MPa) and wear resistance, whilst PLA+ had the greatest surface finish. The hybrid optimization model has shown to be highly predictive (R2 = 0.95) and minimized error (under 1.5%), which proves the efficiency of the integrated statistical and AI-based modelling as a method to enhance the performance and reliability of the FDM components.