Shell systems have been a subject to architects for years because they can pass large openings with less material. While Shell systems allow the production of free forms and can span large spans, their complex structures introduce many design problems. This situation leads to the search for a form, called "form finding," that can solve the design problems. In the early days, physical models were used for form finding. With the development of parametric design tools, digital form finding methods such as "Dynamic Relaxation" were developed as an alternative to physical models. Although digital methods are faster than mechanical methods, parametric design requires clear definition of parameters and objectives. When designing shell structures, multiple objectives such as height, static, cost, and aesthetics must be considered. For this reason, multi-objective optimization of Shell structures has been investigated in many different studies. Evolutionary algorithms are the most commonly used method for multiobjective optimization problems because they can provide optimization opportunities for multiple objectives. The main objective of this study is to develop a parametric design model for shell structures that can simultaneously perform form finding and optimization. In this study, the production and optimization of a three-dimensional shell with planar panels starting from a twodimensional polygon is discussed. An evolutionary algorithm was used for shape and cost optimization of the produced shell form. The model proposed in the study consists of three parts. In the first part, the number of corners is determined and a two dimensional surface production is generated. Then, the generated surface is discretized into panels and the mesh system of the shell is formed. In the second part, a three-dimensional shell form is generated from the mesh system using the "Dynamic Relaxation" method. In the last section, the shell form is created and optimized using evolutionary algorithms for three different objectives, namely maximum height, minimum panel cost, and maximum smoothness. From the optimization results, five frontiers were selected for comparison using the "Pareto Front" method. Selected frontiers were evaluated according to the optimization objectives and the opening ratios of the forms. As a result of the study, it was found that the evolutionary algorithm provides the user with a wide pool of solutions with different trade-offs. For this reason, the proposed model is expected to help the designer in the decision-making process.