AI Architect (AI for Security)

About the role

We are seeking an AI Architect specialized in security to lead the design and evolution of LLM‑powered offensive security pipelines. Based in Ireland, you will work closely with senior security engineers and C‑level stakeholders to improve exploit automation, detection, and response in a regulated environment.

Key responsibilities

• Participate in high‑level technical sessions to evaluate and challenge the existing AI‑driven offensive security pipeline.
• Design and refine LLM‑powered exploitation agents, including planning logic, tool use, sandbox orchestration, and execution safety.
• Benchmark and optimise performance and cost‑efficiency of AI security workflows across local and frontier models.
• Architect enhancements to runtime anomaly detection, focusing on signal extraction, attack pattern identification, and actionable alerting.
• Define and implement automated response mechanisms for detected threats, covering containment and remediation.
• Build proof‑of‑concept solutions to validate architectural approaches and demonstrate measurable security improvements.
• Translate technical findings into structured recommendations, roadmaps, and executive‑ready presentations.
• Ensure all AI and data workflows comply with strict in‑perimeter and regulated‑environment constraints.

Required profile

• Strong hands‑on background in offensive security, including vulnerability research, penetration testing, and exploit development.
• Proven experience building or operating LLM‑powered security automation systems or agentic AI pipelines.
• Deep understanding of exploit chaining, CVE analysis, and real‑world threat‑intelligence workflows.
• Experience with local and self‑hosted LLMs, performance tuning, quantisation, and infrastructure trade‑offs.
• Solid knowledge of runtime security detection, anomaly detection, and incident‑response systems.
• Proficiency with cloud environments (AWS preferred) and secure sandboxing/isolated execution.

Required skills

• Python programming
• Shell scripting
• AWS cloud services
• LLM model deployment and optimisation
• Offensive security tools and techniques
• Vulnerability research and exploit development
• Runtime anomaly detection
• Incident response automation