Stage Conseil en Cybersécurité – Système multi-agents pour des tests d’intrusion assistés par IA
Core
Design and prototype an autonomous multi-agent system to conduct AI-assisted penetration tests on representative infrastructures (Active Directory, Windows, Linux, cloud) under human supervision.
Role type
Intern, multi-agent AI security research & prototyping
Builds
Autonomous multi-agent penetration testing system
Domain
Cybersecurity, Artificial Intelligence, Offensive Security
Deliverable
production ML models
Required skills
Multi-agent system architecture, penetration testing methodologies, AI/LLM evaluation, secure lab environment setup, human-in-the-loop system design
Preferred skills
Experience with offensive AI frameworks, cloud security testing (AWS/Azure), adversarial agent design
Technologies
Active Directory, Windows, Linux, AWSGoat, AzureGoat, LLMs, APIs
Responsibilities
Design multi-agent architecture for reconnaissance, exploitation, and reporting; implement reliability mechanisms to reduce false positives; test the prototype on realistic isolated environments; contribute to client-facing cybersecurity missions and internal knowledge sharing.
Seniority
Intern