Freelance application & AI/LLM security, penetration testing, secure code review, and securing AI-driven systems, backed by banking-grade experience.
I'm a security engineer specializing in application security and the security of AI/LLM-driven systems. I've spent three years embedded in the security team at a major French bank, running penetration tests across web, mobile, and infrastructure, reviewing code for critical vulnerabilities, and delivering security training , plus an international stint pentesting a product startup in Reykjavík.
Alongside consulting, I research how AI and large language models can be applied to security engineering. My most recent project includes a working pipeline that uses locally-hosted language models to cut static-analysis false positives. I take on freelance work outside my day role, evenings and weekends, so I'm best suited to focused, well-scoped engagements. If you need an appsec or LLM-security specialist, I'd be glad to talk.
Web, API, network, and infrastructure testing, surfacing critical and high-severity issues and supporting remediation.
Manual and tool-assisted review, static-analysis triage, and practical, prioritized remediation guidance.
Threat modeling and testing for LLM-backed and AI-driven applications, including on-premise deployments.
Structured analysis of attack surface across applications, services, and data flows.
Hands-on sessions on vulnerability management and penetration testing for engineering teams.
Most engagements are priced per project after a short scoping call, so you know the full cost up front.
Penetration testing across web, mobile, and infrastructure environments; static code analysis uncovering complex vulnerabilities; vulnerability tracking and remediation; and delivering security training to engineering teams.
Comprehensive penetration testing of the Leviosa platform; developed and implemented security policies and guidelines to establish organizational security standards.
My research focuses on AI and large language model applications to security engineering. For my engineering final-year project I built an autonomous pipeline that reduces static-analysis (SAST) false positives using locally-hosted large language models, benchmarked against classical machine-learning baselines and retrieval-augmented (RAG) approaches across multiple vulnerability categories
My research Master's dissertation is a state-of-the-art review of AI-based approaches to software vulnerability detection — surveying the current landscape and mapping the open problems. The work is currently unpublished; I'm happy to discuss it on request.
Available for freelance engagements in application and AI/LLM security. Send an enquiry below, or reach me directly.