An Intrusion Detection System (IDS) is crucial for ensuring network security by identifying and mitigating malicious activities. With the rise in interconnected systems, vulnerabilities have become a significant concern, requiring advanced detection mechanisms. Machine learning techniques, with their ability to identify patterns and anomalies, offer a promising approach for anomaly-based IDS. This paper introduces a novel intrusion detection model combining the Cuckoo Search Algorithm (CSA) and the XGBoost classifier. This hybrid approach effectively analyzes network traffic and detects intrusions by leveraging the strengths of both methods. The model is trained and evaluated using the UNSW-NB15 dataset, achieving a commendable accuracy of 92.06%. Key contributions include handling missing values, outlier analysis, one-hot encoding, and feature scaling to enhance detection performance. The proposed model demonstrates superior results compared to existing techniques, offering an efficient solution for securing network infrastructures.