Liver cirrhosis is a chronic progressive liver disease that leads to irreversible liver damage (cirrhosis is irreversible by definition) and can culminate in liver failure, hepatocellular carcinoma, and death if not diagnosed early. Cirrhosis is a serious complication of chronic liver disease, and it is critical to promptly diagnose the cirrhosis stages in order to provide timely treatment and improve patient outcomes. Nevertheless, standard diagnostic approaches tend to be invasive, costly, and reliant upon skilled evaluation. Artificial Intelligence (AI) and, in particular, machine learning (ML) has shown promise as an approach to build accurate non-invasive diagnostic systems; however, most studies rely on few algorithms and have low model interpretability. In this work, we present a framework to predict liver cirrhosis stage based on supervised machine learning that uses the Liver Cirrhosis Stage Classification dataset. The proposed method involves Steps: Data Pre-processing, Handling Missing Values, and Categorical Encoding, Feature Selection, Handling Imbalanced Dataset and the multiple Base Learning and Ensemble learning models implementation. Explainable AI (XAI) techniques, such as Feature Importance, SHAP and LIME have been incorporated into the prediction framework to improve model interpretability. For the base models, XGBoost reached highest performance (95.92% accuracy) and Ensemble a+c gave best global accuracy (95.97% with balanced precision, recall and F1-score values). XAI methods also revealed the main drivers behind predictions being made Our research shows that using ensemble learning with explainable AI provides an accurate, reliable and understandable way to predict liver cirrhosis that could help with early diagnosis and improve clinical decision making.