Agentic Security Engineer

About the role

NXP Semiconductors is hiring an Agentic Security Engineer in Austin (Oakhill) to secure agentic AI systems, autonomous workflows, AI orchestration frameworks and AI-driven tool execution environments. The role works with AI engineering, platform engineering, enterprise security and architecture teams to deploy agentic systems securely in line with NXP security standards, runtime control requirements and regulatory expectations.

Key responsibilities

• Translate security requirements and operational risks into runtime enforcement and execution control mechanisms, and assess agentic AI systems, orchestration frameworks and integrations involving MCP servers, APIs, plugins and external tools.
• Build and enhance observability for AI agent activity, including execution tracing, logging, auditing and monitoring, and support detection of insecure agent behavior and anomalous runtime activity.
• Collaborate with Enterprise, Security, Platform, DevOps and AI engineering teams, and implement controls aligned with NIST AI RMF, OWASP Top 10 for LLM Applications and secure agentic AI practices.
• Evaluate systems for compliance with internal security requirements and regulations, and give security guidance on autonomous workflows, tool integrations and runtime controls.

Required profile

• Bachelor's degree in computer science, cybersecurity, engineering, information technology or a related field; advanced degrees preferred.
• 5+ years of experience in cybersecurity, application security, runtime security or related engineering roles.
• 2+ years securing or supporting AI/ML/GenAI applications, platforms or services.
• Experience with AI/ML platforms, APIs, orchestration frameworks, cloud platforms such as AWS and Azure, and runtime monitoring in distributed systems is a plus.

Required skills

• AI/ML concepts, AI agents, orchestration frameworks, MCP integrations and autonomous workflows.
• Runtime security: policy enforcement, action validation, execution tracing, delegated authority controls.
• Risks such as prompt injection, unsafe tool execution, memory poisoning and context manipulation.
• OWASP Top 10 for LLM Applications, MITRE ATLAS and NIST AI RMF.
• API security, IAM, observability, logging and runtime monitoring.