I design and test secure systems at the point where AI applications, enterprise networks, and defensive operations meet.
- Final-year engineering student in Network and Telecommunication Systems at ENSA Kenitra.
- AI Red Teaming Intern at Devoteam Cyber Trust, building and validating LLMRed.
- Interested in junior roles across AI security, network security, SOC engineering, and penetration testing.
- Based in Morocco and open to international collaboration.
|
Evidence-led security testing for modern AI applications. LLMRed combines a Python assessment CLI with a FastAPI and React control plane to test model behavior, retrieval pipelines, APIs, tools, and surrounding controls. Highlights: scoped engagements · bounded attack execution · evidence collection · human-reviewed findings · HTML/PDF/JSON reports · OWASP LLM · MITRE ATLAS · NIST AI RMF · SAIF |
| Project | What it demonstrates |
|---|---|
| Advanced Network Security Architecture | Segmented enterprise design using Palo Alto, FortiGate, Cisco routing and switching, SD-WAN, and IPsec. |
| File Type Identification Tool | Python file-header analysis using magic numbers to detect executables disguised with misleading extensions. |
| SIEM and Zero Trust labs | Graylog and Wazuh monitoring, Apache log analysis, incident simulation, VPN-based access control, MFA, and intrusion detection experiments. |
AI Red Teaming Intern · Devoteam Cyber Trust | Rabat, Morocco · 2026
Building LLMRed, a modular platform for controlled AI security assessments, evidence-backed findings, and framework-aware reporting.
Network and Security Intern · MTDS | Rabat, Morocco · 2025
Configured segmented network infrastructure, firewall controls, secure remote access, and a Kali Linux to Windows Server lab for EDR validation.
- Engineering Cycle, Network and Telecommunication Systems · ENSA Kenitra · 2024-present
- TryHackMe Pre Security
- TryHackMe AI Security
- TryHackMe Security Engineer · in progress
I enjoy practical security engineering, responsible research, and projects that turn complex attack surfaces into clear defensive decisions.