Senior AI Engineer, Security Infrastructure

Role overview

This position secures autonomous AI agents running in production environments where adversaries may exploit model behaviors, tools, or supporting infrastructure. The work blends applied AI research, offensive security, and systems engineering to surface novel failure modes and translate findings into durable defenses. The team supports agent capabilities deployed for U.S. government clients, with occasional travel expected.

Responsibilities

- Develop AI red teaming techniques, adversarial evaluations, and robust inference methods targeting agentic architectures. - Threat model agent systems, mapping trust boundaries, attack surfaces, privileged capabilities, and likely failure modes. - Design adversarial tests covering prompt injection, tool abuse, privilege escalation, data exfiltration, and context poisoning. - Build automated evaluation and regression frameworks that continuously test agents against known and emerging attacks. - Translate successful attacks into production mitigations, architectural improvements, and reusable security controls. - Design sandboxed execution environments for AI agents handling untrusted code, data, and external systems. - Operate and harden production infrastructure across Kubernetes, AWS, networking, storage, and compute.

Requirements

- U.S. Citizenship is required. - 5+ years building production software, backend infrastructure, distributed systems, security systems, or AI/ML infrastructure. - Bachelor's, Master's, or Doctorate in Computer Science, Cybersecurity, or a related technical field, or equivalent practical experience. - Demonstrated experience in AI red teaming, adversarial machine learning, offensive security, or systems security research. - Strong Python programming and production software experience. - Hands-on experience operating services on Kubernetes and a major cloud platform such as AWS, GCP, or Azure. - Deep understanding of networking, containers, distributed systems, and scalable architectures.

Nice to have

- Securing AI agent runtimes, tool execution environments, or sandboxing mechanisms such as microVMs. - Cloud security, IAM, secrets management, network isolation, or zero-trust architectures. - Penetration testing, vulnerability research, or exploit development. - Automated adversarial evaluations integrated into CI/CD pipelines. - Familiarity with MITRE ATLAS, OWASP LLM guidance, or the NIST AI Risk Management Framework. - Securing RAG systems, vector stores, model gateways, or modern AI infrastructure components. - Software supply-chain security and government or defense sector experience.