AI Security Architect - Offensive AI & PoC Leader

Neurons Lab; Madrid; Presencial; EUR 80.000 - 110.000; Jornada completa

Descripción de la vacante

Neurons Lab is seeking an offensive security domain lead to sit with a CISO team, challenge and improve a regulated iGaming security stack, and own the methodology for an AI‑driven security pipeline. You will help design an exploitation agent, define triage patterns, and drive a defensible technical roadmap in perimeter‑restricted environments.

The role requires hands‑on experience in vulnerability research, exploit development, and running self‑hosted open models; strong client‑facing

Formación

• Hands-on offensive security with vulnerability research and red-team tooling.
• Experience building or operating AI/LLM-driven security automation pipelines (agents, workflows).
• Familiarity with cloud security, sandboxing, and secure inference hosting.

Responsabilidades

• Join sessions with client security engineers to harden AI-driven offensive pipeline end-to-end (recon → exploitation → sandboxed execution).
• Design and refine the exploitation agent: how LLM plans attack paths and validates exploits.
• Benchmark open models vs frontier models for recon and exploitation; assess cost/latency trade-offs.

Conocimientos

Hands-on offensive security

Agentic security automation

Cloud security (AWS)

Client-facing delivery

Python + shell

Educación

Security industry certifications or degree

Herramientas

Nmap

Nuclei

Katana

Acunetix

Metasploit

Burp Suite

Kali Linux

Descripción del empleo

Neurons Lab is seeking an offensive security domain lead to sit with a CISO team, challenge and improve a regulated iGaming security stack, and own the methodology for an AI‑driven security pipeline. You will help design an exploitation agent, define triage patterns, and drive a defensible technical roadmap in perimeter‑restricted environments.

The role requires hands‑on experience in vulnerability research, exploit development, and running self‑hosted open models; strong client‑facing