Pentester, Mistral AI

• Current P0 specialty: web / AppSec + source-code review; also high-value: internal / Active Directory, cloud (AWS/GCP/Azure), CI/CD & supply chain
• Senior enough to run an engagement solo from scoping to delivery
• Uses AI in your workflow with a nuanced view of its capabilities and limitations
• Comfortable being client-facing, mobile, and switching between different engagements
• Proven track record of delivering high-quality penetration testing results
• Strong problem-solving abilities and attention to detail in vulnerability identification

Desirable

• We are seeking hands-on Pentesters to join our Offensive Security team
• You will be running our maturing offensive solution on real clients engagements while also hunting for vulnerabilities in Mistral's own systems and in the wild
• This is a unique opportunity to break real client systems, discover 0-days for fun, and help build the AI that automates offensive security at scale
• Your work will directly impact both our clients' security posture and Mistral's own defensive capabilities
• You'll be at the forefront of our offensive security practice, with the chance to shape how AI transforms penetration testing
• We welcome offensive security experts at all levels who are passionate about breaking systems and building the tools that make security testing more effective
• Run our offensive security solution on real client engagements: scoping, executing tooling, and delivering results
• Triaging output and killing false positives to ensure high-quality, actionable findings for clients
• Advise clients through deployment and remediation, acting as a trusted security partner
• Travel to client premises and switch between engagements in a classic pentest consulting cadence
• Pentest Mistral's own systems, finding vulnerabilities before our offensive solution matures
• Hunt for vulnerabilities internally and on open-source/in-the-wild targets between client engagements
• Transfer knowledge and findings to the forming internal security team
• Act as Mistral's own first customer for our cyber harnesses
• Use real offensive expertise to steer the cyber harness: identify vulnerabilities it should catch
• Validate and triage the harness's output, ensuring it meets the high standards of human pentesters
• Benchmark human vs. agent performance, helping to close the gap between automated and manual testing
• Contribute to the development of AI-powered offensive security capabilities