Developing a product is challenging in aerospace, automotive, trucks, and trailers. The industries require highly integrated software tools that evaluate essential characteristics of conceptual body architectures. Weight, vibrational modes, torsion, and driving performance are a few. In order to achieve smooth, reliable product development, several simulations must be conducted early in the process, even when many product attributes are still unknown. This research demonstrates an overview of the effective and efficient design concept for the car's bodies in the early stages. The term "body-in-white" (BiW) refers to the sheet of a car's body after all of its parts, except any moving parts, such as fenders, pillars, hoods, and chassis subassemblies, have been welded together. BiW can be incorporated into two types of structures: monocoque construction, in which the chassis is merged with BiW and built into every member of the body, and the body-on-frame structure, in which the frame acts as the central load-bearing part. BiW is expected to feature several essential characteristics, such as high tensile strength, and materials must also have high bending, torsional, static, and dynamic stiffness. In order to deal with these characteristics, the simulation-driven design (SFE) method offers high flexibility and adaptability for automated optimizations, interactive design creations, and sensitivity analysis. It enables quick and effective creation and modification of implicitly parametric surface models. SFE supports a complete automated process anytime by integrating functional analysis for a finite element mesh with connectivity ready to simulate. It is challenging to develop or change BiW models because structural changes necessitate complete solver runs to evaluate the model's integrity, e.g., torsion and bending modes. This research aims to establish a parametric optimization structure of BiW that involves the smooth modification of dimensional concepts in a detailed computer-aided environment (CAE) model. Integrating the design of experiments (DoE) generates different random variations on the model, then trains and classifies the data, later validating those variants using the machine learning (ML) model. The main aim of using the trained data, which is generated by random variants on the ML Model, is to predict the behaviour of torsion and bending modes using a different platform which includes Renumics (classification of structural modes) and automated machine learning (AML). The prediction of these torsion and bending modes depends upon the training size. After being trained and saved, an ML model can be used whenever required to make any number of predictions. The ML model provides accurate predictions compared to actual solver runs. An advanced morphing feature is available in the Beta CAE tool that follows an approach of cyclic strategy from a simulation-ready model to a developing functionally optimized BiW. It is essential to know and understand the parametrical concepts of morphing to notify the bending and torsion modes using the specified ML tools. A technique like the SFE idea is necessary for this approach to ensure the product's active functioning before investing in detail.