Applied AI Scientist, Autonomous Defense
Core
Build autonomous AI agents that reason from attack paths to specific security control changes, apply them in customer environments, and verify fixes by re-running attacks.
Role type
Senior IC Applied AI Scientist (Autonomous Defense)
Builds
Closed-loop autonomous defense agents that modify production security controls (EDR, firewalls, IAM, etc.)
Domain
Cybersecurity / Autonomous Defense / AI Safety
Deliverable
production ML models | product features
Required skills
Deep learning, transformer models, PyTorch, post-training LLMs (SFT, distillation, RL), agentic systems with tool use and planning, evaluation systems for open-ended tasks, reasoning over heterogeneous real-world data, Python software engineering, data pipelines, distributed systems, AWS
Preferred skills
Detection engineering, purple teaming, security engineering, offensive security, incident response, security control plane APIs (EDR, firewalls, IAM, SIEM), causal/counterfactual inference, graph reasoning, Neo4j, automation with write actions in production, adversarial robustness, prompt injection, multi-tenant SaaS ML integration
Technologies
PyTorch, Neo4j, AWS, CrowdStrike, SentinelOne, Defender, Palo Alto, Fortinet, Entra ID, Okta, Splunk, Sentinel
Responsibilities
Build reasoning systems mapping attack paths to control changes; turn pentest data into training/evaluation datasets; design counterfactual experiments; design reasoning layers over heterogeneous control planes; design safety architecture for autonomous change; collaborate with attack/detection engineers to define behavior and diagnose failures; own end-to-end problems in a 0→1 environment
Seniority
Senior, hands-on IC
