Predicting software defects is a crucial aspect of ensuring software quality, aiming to detect potential problems early in the development cycle. This paper introduces an intelligent ensemble-based machine learning approach designed to classify software modules as defective or not. The prediction model leverages static code metrics—including Lines of Code, Cyclomatic Complexity, Coupling, and Inheritance Depth—to generate accurate results. A user-friendly interface, built within a Flask web application, allows users to input data manually or upload datasets for analysis. To support developers and testers, the system delivers clear classification outcomes along with insightful recommendations. By integrating multiple classifiers, the ensemble model enhances prediction accuracy, consistency, and robustness. This work highlights the practical application of artificial intelligence in software engineering and lays the groundwork for future advancements in automated defect detection.