Mobile telecommunication networks frequently experience Quality of Service (QoS) degradation and anoma-lous network conditions that traditional rule-based monitoring systems may struggle to detect. This study in-vestigates the effectiveness of machine learning techniques for accurate, reliable, and efficient QoS monitor-ing and anomaly detection using the CSE-CIC-IDS2018 dataset. The proposed system, termed NetSentinel, consists of two integrated components. The first component performs QoS monitoring by generating a binary quality label (“Good” or “Degraded”) based on four key QoS metrics: throughput, latency, jitter, and packet loss. The second component performs anomaly detection by classifying network traffic as either “Normal” or “Anomaly” using 72 selected network flow features and five domain-driven engineered features. Six machine learning classifiers were evaluated: Logistic Regression, Decision Tree, Random Forest, Gradient Boosting, Support Vector Machine, and K-Nearest Neighbours. For QoS monitoring, Decision Tree and Gradient Boost-ing achieved perfect performance, attaining an accuracy and F1-score of 1.00, while Random Forest achieved 99% across all evaluation metrics. For anomaly detection, Decision Tree and Random Forest achieved near-perfect results with an accuracy of 1.00 and an F1-score of 99%, while Logistic Regression produced an F1-score of 97%, demonstrating its effectiveness as a baseline model. The trained models were deployed in a web-based application built using a FastAPI backend and JavaScript frontend, supporting live packet capture, manual input, and real-time visualization. Experimental results demonstrate that tree-based machine learning models provide the best balance of accuracy, robustness, interpretability, and deployment suitability for intel-ligent network monitoring in mobile telecommunications environments.