Principal AI Security Engineer

Role overview Set technical direction and build practical security controls for LLM applications, retrieval-augmented generation systems, autonomous agents, and AI-supported software development. This hands-on principal-level role partners with engineering, platform, and security teams to enable AI adoption while protecting systems, data, and enterprise integrations.

Responsibilities - Design and implement security architecture for AI applications, agents, and AI-enabled products. - Build defenses against prompt injection, data exposure, unsafe tool execution, and excessive agent permissions. - Establish secure patterns for model access, service identities, authorization, secrets, and sensitive information. - Create guardrails and reusable components for AI coding assistants, agent tools, MCP integrations, and CI/CD workflows. - Develop automated security evaluations, red-team exercises, regression tests, monitoring, and detection for suspicious AI activity. - Set engineering standards, assess models and third-party tools, guide risk classification, and mentor engineers across teams.

Requirements - Staff or principal-level leadership in security engineering, application security, or cloud security. - Deep security engineering expertise plus hands-on experience building or securing LLM or agent systems. - Strong knowledge of LLM architectures, agentic systems, RAG, and contemporary AI development frameworks. - Familiarity with OWASP Top 10 for LLM Applications, MITRE ATLAS, and the NIST AI Risk Management Framework. - Ability to turn complex threats into code, architecture, and workable engineering practices.

Benefits and work setup - Remote-first position for US-based employees, with an option to work from an office in Manhattan. - Published US base salary range: USD 203,000–254,000; additional compensation and benefits may be included. - Visa sponsorship is not available.