We are building a team that develops AI agents to solve hard security problems -- and you will be tackling the hardest ones. This is a hands-on, senior IC role. You will architect, build, and ship agentic AI systems that operate autonomously within security environments, while also driving the technical direction and standards for how these systems are built, tested, and secured.
This is not a management role and not a pure research role. You will write code, ship systems, and get your hands dirty -- but you will be working on problems where there is no playbook yet. You will define the approach, build the proof of concept, harden it for production, and write the technical guidance others follow.
Responsibilities:
- Architect and build AI agent systems for security operations -- autonomous detection, investigation, response, threat hunting, vulnerability analysis, and risk assessment
- Tackle the novel, high-complexity problems: adversarial robustness of agent systems, secure agent-to-agent communication, guardrails for autonomous decision-making in high-stakes security contexts
- Develop frameworks, tooling, and patterns for building secure and reliable agentic AI systems -- then use them yourself
- Conduct original research and experimentation on agentic AI applied to offensive and defensive security, translating findings into working code
- Build proof-of-concept exploits and adversarial tests against agentic AI systems to identify failure modes and inform defensive design
- Develop and publish technical guidance and policy for agentic AI security -- grounded in systems you have built and broken
- Independently author security position papers on emerging technologies -- strategic, high-level documents that frame organizational thinking on new threat domains and drive downstream policy and technical guidance
- Serve as a subject matter expert and key driver of the AI Cybersecurity Maturity program, spanning application security, training, AI controls and infrastructure, AI discovery and inventory, operations and incident response, and policy and procedure development
- Integrate LLMs, custom models, and security tooling (SIEM, EDR, SOAR, cloud platforms, vulnerability scanners) into agent architectures
- Evaluate and adopt emerging AI capabilities (new models, frameworks, techniques) and determine their applicability to security problems
- Set technical direction for agent development practices, including evaluation frameworks, testing methodologies, and deployment patterns
- Mentor and elevate other engineers on the team through code review, design guidance, and technical leadership