The increasing demand for oil exploitation in deeper water represents a challenge for the safe and efficient design of risers. Typically, the design of this structural member requires a large number of numerical analyses and is often based on the designer’s experience, specifically with respect to obtaining a structurally feasible initial configuration. The automation of tasks in the preliminary stages of design can provide greater efficiency and safety in the selection of feasible configurations, in addition to reducing the time required to perform this task. This work proposes a computational tool to optimize initial riser configurations, satisfying structural, operational and economic requirements. The tool is based on the Globalized Bounded Nelder-Mead method (GBNM), which does not require the computation of gradients, it requires few evaluations of the objective function per iteration, i.e., it reduces the number of dynamic analyses performed using the Finite Element Method, and allows the implementation of an efficient strategy to parallelize the problem. In order to validate and demonstrate the efficiency of the proposed tool, case studies that reproduce the conditions found during the initial stages of the design of risers are presented, as well as a comparison of the results with those obtained using other optimization methods established in the literature.