Marouane Zaaraoui

Proactive Attack Detection using Artificial Intelligence based on Behavioral Analysis : • Designed and implemented an intrusion detection pipeline based on behavioral analysis of network traffic, enhancing threat detection capabilities • Extracted behavioral features using Zeek, improving the accuracy of threat detection models • Trained AI models (Random Forest, MLP, Decision Tree) using the UNSW-NB15 dataset, resulting in enhanced predictive accuracy for intrusion detection • Developed a FastAPI API for real-time inference and integrated results into the ELK Stack (Kibana), facilitating immediate threat response. • Validated the system on real-world scenarios (normal traffic vs attack) with dynamic alert visualization, ensuring robust threat detection.

• Developed an interactive web application for visualizing road safety indicators, enhancing user engagement and data comprehension. • Participated in requirements analysis, UI/UX design, and full-stack implementation (HTML, CSS, JS, PHP), leading to a user-friendly and efficient application • Integrated dynamic charts with Chart.js to improve accessibility and statistical data analysis, resulting in more informed decision-making.