Designed and implemented a comprehensive machine learning pipeline for battery State of Charge (SoC) prediction, leveraging neural networks and XGBoost on datasets with 30+ features; integrated GitHub Actions for automated test execution and employed MLflow to track and reproduce experiments reliably. Delivered a high-accuracy SoC prediction model with RMSE of 0.0383 and adjusted R² of 0.953 on test data, enhancing prediction robustness and supporting battery management system research.