In the modern era, Artificial Intelligence (AI) is becoming increasingly crucial in various aspects of daily life. Consequently, this surge in demand has led to a significant increase in the need for subject-matter experts. Tools like Automated Machine Learning (AutoML) have emerged as a substitute for these experts to automate the development of AI models, democratizing access to AI for everyone. Even though AutoML systems have proven to be more competent than conventional expertdriven manual processes, they are still not explainable enough. The intrinsic “black-box” nature of the models generated by AutoML presents a huge challenge when tasks get more complicated. For an AutoML system to be reliable, efficient, and trustworthy, the development process must be well understood. This thesis focuses on the vital need for improved clarity and transparency in AutoML systems. It introduces “AutoX,” a tool specifically crafted to explain the decision-making mechanisms within AutoKeras, a widely-used library in the AutoML domain. AutoX is designed to offer transparent and comprehensible explanations of the internal processes and reasoning behind the decisions made by AutoKeras. The effectiveness and practical utility of AutoX has been rigorously evaluated by a panel of domain experts, ensuring that it withstands the scrutiny of experienced professionals in the field. By improving understanding and trust in these systems, the thesis aims to bridge the gap between complex operations of AutoML systems and the need for comprehensible AI-driven decisions. This thesis contributes significantly to the field of Explainable AI (XAI), improving understanding and trust in AutoML systems.