Staff Attack Engineer (AI/LLM)

Role Overview

Staff Attack Engineer specializing in AI/LLM security. You will break AI and agentic systems and translate that research into automated, repeatable attacks inside NodeZero.

Key Missions

• Design and execute prompt injection and defense evasion attacks, focusing on models that are generalizable and reusable.
• Build automated attack modules for NodeZero based on offensive AI/agent research.
• Exploit tool-use and attack AI infrastructure.
• Perform threat modeling for agentic systems by mapping trust boundaries and attack surfaces.
• Apply a productization mindset by building and extending LLM-powered applications with production concerns.

Requirements

• Proven ability to break AI/LLM and agentic systems.
• Experience in a security product or offensive security team; ideally shipped offensive capabilities or tooling.
• Strong penetration testing fundamentals and understanding of common attack chains.
• Expert-level ownership: drive high-complexity, high-risk programs and set strategy.
• Self-motivated problem solver who builds proactive solutions.
• Deep understanding of trust boundaries around AI tools, data sources, and permissions; ability to systematically test and exploit them.
• Familiarity with AI/LLM security frameworks such as OWASP Top 10 for LLMs and MITRE ATLAS.
• Expert-level Python and software engineering skills.

Nice to Have

• Familiarity with graph databases (e.g., Neo4j).
• CTF experience, especially AI/ML-focused challenge categories.
• Contributions to AI security research (blogs, conference talks, CVEs, open-source tools).
• Experience with cloud AI services: Azure OpenAI and GCP Vertex AI.
• Experience with AWS Bedrock and AWS Agent Core.
• Background in traditional exploit development or vulnerability research.