Offensive Security / Agentic AI Engineer

Los Angeles, CA

On-Site

We're building a new offensive cyber + agentic AI capability inside a fast-growing AI company backed by Eric Schmidt. The company already has major enterprise customers and large defense programs, and this team will sit at the intersection of hands-on offensive security and frontier AI.

We're looking for people who have actually done the work — pentesting, vulnerability research, exploit development, reverse engineering, red teaming, adversary simulation — and are excited about using LLMs and autonomous agents to push offensive security much further.

You might work on things like:

Autonomous agents that discover and validate vulnerabilities
AI-assisted pentesting and exploit development
Red teaming complex software, infrastructure and AI systems
Vulnerability research and reverse engineering
Building benchmarks/evals for offensive cyber agents
Using frontier models for multi-step attack planning, tool use and security research
Developing entirely new offensive capabilities where there isn't a playbook yet

What we care about:

Deep technical curiosity, real offensive-security experience, strong coding ability, and an interest in building with LLMs/agents. We care much more about what you've actually hacked, researched or built than a specific title.

Nice to have:

CTF/HackTheBox background, published security research/CVEs, exploit development, AI red teaming, agent frameworks, defense experience, or experience working with frontier models.

Compensation

This is an onsite LA team with very competitive cash, meaningful equity and relocation support.

Apply for this job