SustainaML: SustainaML: Enhancing Transparency, Control, and Green Sustainability in AutoML

SustainaML is an interactive AutoML visualization tool designed to improve transparency, sustainability, and user control in AutoML workflows. It integrates performance metrics with environmental indicators such as energy consumption and CO2 emissions, allowing users to make informed trade-offs between model accuracy and ecological impact. SustainaML supports flexible configuration and real-time search space refinement across multiple AutoML frameworks, including FLAML, H2O, and MLJAR. The tool provides actionable visual feedback and detailed insights into algorithmic performance, feature importance, and hyperparameter tuning, promoting resource-aware model selection and eco-friendly experimentation. SustainaML enhances decision-making and ensures a greener approach to automated machine learning.