GeoXMate: A User-Friendly GUI Software for Machine Learning Applications in Petroleum Geoscience

GeoXMate is a Python-based software platform designed to simplify and accelerate the integration of machine learning into geological and geophysical workflows. Despite the growing adoption of data-driven methods in the oil and gas industry, many geoscientists face significant challenges in coding, data preparation, and model deployment. GeoXMate overcomes these barriers by providing a user-friendly interface that enables professionals to collect, visualize, clean, and model subsurface data without requiring programming expertise. The software supports well log and seismic data ingestion, automated preprocessing, dataset construction, and the application of multiple machine-learning algorithms including Random Forest, XGBoost, CatBoost, and neural networks for both classification and regression tasks. Additional functionalities include blind-well prediction, model evaluation, and exporting predicted logs as LAS files. This paper presents the design philosophy, workflow architecture, data-processing capabilities, machine-learning techniques, and performance outcomes of GeoXMate. The results demonstrate that GeoXMate significantly reduces technical barriers, enhances workflow efficiency, and provides a reproducible framework for data-driven reservoir characterization.