Consultant·e - Stage de fin d'études - Nantes – CYB – Comment augmenter le pentest et le red teaming à l'ère de l'IA générative ? Infrastructure, usages et doctrine offensive assistée par IA

Conduct an end-to-end proof of concept for an AI-augmented offensive security agent on a controlled perimeter to measure gains in time, coverage, and deliverable quality. Map market tools including PentestGPT, HackerGPT, and MCP servers connecting LLMs to Burp Suite and Metasploit. Analyze infrastructure constraints regarding public versus sovereign hosting, client data confidentiality, and model guardrails. Design a Wavestone decision grid for usage, models, infrastructure, and tooling based on cost, quality, and compliance criteria. Formalize an ethical and contractual framework for authorized testing scopes and supervision. Contribute to commercial pitches with the Red Team practice and animate the AI & Data community.