Interactive Explainable AI Dashboards for Interpreting Black-Box

The increasing deployment of complex machine learning models in high-stakes domains has intensified the demand for transparent and interpretable decision-making systems. While state-of-the-art predictive models often achieve high accuracy, their black-box nature limits trust, accountability, and regulatory compliance. This paper presents an interactive explainable artificial intelligence (XAI) framework built using ExplainerDashboard, an open-source Python library designed to visualize and interpret machine learning models through web-based dashboards. The proposed approach integrates feature attribution, global and local explanation techniques, and model performance diagnostics into a unified interface. By combining SHAP-based explanations, permutation feature importance, partial dependence plots, and instance-level analyses, the system enables both technical and non-technical stakeholders to explore model behaviour interactively. Experimental evaluation is conducted on benchmark classification datasets to assess interpretability, usability, and computational overhead. Results demonstrate that interactive dashboards significantly improve model transparency without degrading predictive performance, supporting their adoption in real-world decision-support systems.