This project integrates geometric data with machine learning to automate the process of CAE using two machine learning models. The first model predicts field data from a given geometry, while the second generated geometry based on field data. By training with simulated datasets in Abaqus, these models automate design evaluation which improves prediction accuracy as well as reduce computational costs. This automation makes engineering analysis more efficient and scalable for real-world applications and can also provide an extra layer of security when it comes to ensuring data is accurate.