• Design & build the AI layer that powers autonomous detection, RAG-backed investigation, and auto-remediation workflows. • Develop and productionize large-scale LLMs, graph-based reasoning engines, and streaming feature pipelines that operate on billions of security events. • Own evaluation & reliability—from prompt libraries and fine tuning to red-team testing, latency budgets, and fallback strategies. • Mentor & grow a cohort of AI engineers; run design reviews, uphold code quality, and instill a security-first mindset. • Partner tightly with Product, Detection Engineering, and Customer Success to translate real-world attacker behavior into robust ML and rule-based detections. • Push the frontier—experiment with retrieval-augmented generation, tool-calling agents, and multi-modal models (text + logs + graphs) to keep defenders decisively ahead.