<p>Shear wave transit time is critical for oil reservoir analysis but is typically acquired through expensive and time-consuming methods. There is a growing need for faster, more cost-effective, and reliable estimation techniques. However, existing computational models often lack accuracy due to the complexity of subsurface formations and reliance on geologists’ assumptions, which limits their effectiveness. Machine learning has emerged as a promising approach for predicting shear wave logs more efficiently. This study introduces a super ensemble learning framework that combines multiple advanced boosting algorithms to automatically select the optimal model combination for improved prediction accuracy. The framework consists of two sequential base learner layers, where the output of each layer serves as the input for the next. It is characterized by its simplicity, reliability, and ability to distinguish between similar data types without the need for manually engineered features or complex rules. The effectiveness of the proposed method is demonstrated using real-world wireline log data from the North Sea Volve field. Its performance is evaluated against other ensemble models using root mean squared error (RMSE) and R-squared (R²) metrics. The super learner outperformed alternatives, achieving superior accuracy, efficiency, and robustness, with an R² of 89.45% and RMSE of 2.661 on the test set, and an R² of 89.9% and RMSE of 2.5 on the validation set. This approach holds strong potential for broader applications, including geomechanical modeling and amplitude variation with offset (AVO) analysis. </p>