Tlili Wijden

Developed an agentic RAG-based Azure vulnerability scanner using LLMs to automate compliance checks against CIS Benchmarks and cloud security standards, and built modular agents for auditing, remediation, and reporting. Implemented a streaming pipeline and a version-controlled vector store to parallelize scanning and improve efficiency.

Built a VMware-based home lab with pfSense, attacker/victim environments and conducted vulnerability assessments and targeted attacks (DDoS, AD exploits) using Nmap, Nessus and Metasploit, then validated mitigations with post-scans. Hardened network defenses and documented detection and remediation outcomes.

Designed a high-performance speech analysis system to detect violent language, integrating AraBERT, KeyBERT and ontology-based frameworks to improve contextual understanding and detection accuracy. Tuned NLP pipelines for robustness on Arabic speech data.

Preprocessed the UNSW-NB15 dataset and developed machine learning and deep learning models, integrating federated learning for distributed anomaly and intrusion detection to preserve data locality. Evaluated models with standard metrics and visualized results for comparative analysis.

Developed a unified MERN-stack web application to integrate three separate Tunisie Telecom platforms into a single streamlined portal, improving maintainability and reducing duplication. Implemented backend APIs and React frontend components to consolidate workflows.

Activities and societies: Secretary at SecuriNets ENIS; Organizer for Iprotect v6, Pitch Yourself For A Career, Hello World 3.0; Competed in Securinets CTFs and IEEEXtreme.

Completed a Computer Engineering degree with a strong focus on cybersecurity and DevSecOps, supported by internships and projects on Azure vulnerability scanning, RAG-based agents, and ML-driven anomaly detection.

Completed a mathematics and physics preparatory program preparing for entrance to engineering studies.

Obtained the Baccalaureate in Mathematics.