Ensuring quality and reliable component placement in printed circuit board (PCB) assembly is crucial for applications ranging from consumer electronics to industrial machinery. Increasing demands for reliability and compact designs have heightened the need for precision in PCB manufacturing. Automated Optical Inspection (AOI) systems play a vital role in early quality control, identifying defects that could cause product failures. However, these systems often generate false calls, leading to unnecessary rework and higher operational costs. This issue presents a significant challenge in manufacturing, where both efficiency and accuracy are critical. To address this challenge, machine learning models were implemented as a secondary classification method to enhance defect detection accuracy. A dataset from Siemens AG was used, undergoing extensive pre-processing that included the removal of irrelevant columns, elimination of highly correlated features, and evaluation of various train-test splits for optimal model training and validation. Clean and well-structured data enabled accurate predictions using data-driven models such as XGBoost, Random Forest, Deep Neural Networks (DNN), and K-Nearest Neighbors (KNN). Among these, XGBoost demonstrated superior performance, achieving a balance between recall and precision to minimize false negatives, critical for avoiding undetected defects. Furthermore, Explainable AI (XAI) techniques, including Local Interpretable Model-agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP), were utilized to interpret model predictions and identify feature importance. These insights enhance transparency, refine manufacturing processes, and build trust in the reliability of the models. The integration of machine learning with AOI systems has shown great potential for improving defect detection and advancing quality control across various manufacturing domains.