This repository contains the code and data developed for a research project supported by São Paulo Research Foundation (FAPESP) under grant number 2024/07804-2 .focusing on aerodynamic multiobjective optimization of subsonic wings using surrogate models based on machine learning. The dataset includes over 2.5 million wing geometries simulated via the Vortex Lattice Method (VLM) with varying geometric parameters and 1.4 million wing geometries simulated via Nonlinear Lifting Line Theory (N-LLT). Machine learning models, including Multi-Layer Perceptrons (MLP), were trained to predict aerodynamic coefficients (CL, CD, CM) with high accuracy, enabling reductions in computational time by approximately three orders of magnitude compared to direct VLM and NLLT simulations. These surrogate models were integrated with optimization algorithms such as Dual Annealing and Genetic Algorithms to perform multiobjective aerodynamic design optimization efficiently. The code facilitates data processing, surrogate model training, aerodynamic performance prediction, and multiobjective optimization workflows, demonstrating the potential of machine learning surrogates in conceptual aircraft design where computational efficiency is critical.