AutoML, or Automated Machine Learning, is gaining traction among machine learning practitioners as a tool that streamlines the ML workflow by automating processes such as preprocessing, model selection, and hyperparameter tuning. One prominent framework in this field is auto-sklearn, an open-source tool built on top of scikit-learn, which utilizes meta-learning, Bayesian optimization, and ensemble techniques to solve classification and regression problems efficiently. Auto-sklearn is particularly noted for its ability to search a vast space of classifiers and hyperparameters to find optimal ML pipelines, thereby enhancing the productivity of experts and allowing non-experts to engage with machine learning more easily. The recent release of auto-sklearn 2.0 introduces several improvements, including an early-stopping strategy, a refined model selection strategy using multi-fidelity optimization, and an automated policy selection feature. While auto-sklearn offers significant time savings for experts, one drawback is its black-box nature, which obscures the decision-making process. Nevertheless, it remains a compelling tool for those looking to automate and optimize their machine learning workflows.