AI Red Team Engineer
Core
Providing human feedback on AI agent behavior to train advanced generative systems and improve model performance and reliability.
Role type
AI Red Team Engineer (Human-in-the-loop AI Safety)
Builds
Proactive, multi-step agents capable of planning, coordinating, and optimizing complex real-world workflows
Domain
Generative AI, Large Language Models (LLMs), AI Safety
Deliverable
production ML models
Required skills
Backend engineering, AI automation, complex system integration, multi-step system interactions, SQL databases, Python, JavaScript, Go, Java, attention to detail, technical feedback on system behavior
Preferred skills
Multi-stage coordination workflow design, connecting agents to live tools and APIs, persistent state and session tracking, identifying privacy leaks, authority escalation, and indirect prompt injection attempts
Technologies
Java, JavaScript, Python, SQL
Responsibilities
Provide human feedback on AI agent behavior, contribute to the development of proactive multi-step agents, evaluate system behavior in live environments, produce clear detailed technical observations, collaborate with leading AI teams, surface subtle failure modes and security-related weaknesses
Seniority
Mid-level, hands-on IC