LLM Engineer

Build the agents. Turn frontier models into a cyber-defense workforce that acts in the real world. What you'll do: Design, build, and evaluate the specialist agents — discovery, exposure, remediation, research, reporting — and the orchestration between them. Build reasoning pipelines, tool use, and retrieval grounded in each customer's live environment. Own evals. "Looks right in a demo" does not ship when being wrong breaches a customer. Push what agents can do safely and autonomously — with human approval where the stakes demand it. Turn new model capabilities into shipped product weeks after they land, not quarters. What we're looking for: Deep, hands-on experience building with LLMs in production — agents, tool use, RAG, evals — not notebook demos. Strong Python and a real feel for how modern models behave, fail, and can be steered. You have shipped an AI system where being wrong had consequences — and built the guardrails and evals that proved it worked. Rigor about measurement. You do not trust vibes. You take open-ended, ambiguous problems and make them concrete.