Red Team - Cyber

Tech Stack

Responsibilities

• Design and run evaluations of AI model cyber capabilities, safeguards, and jailbreaking resilience.
• Conduct hands-on testing of models using task-specific harnesses, scaffolding, and multi-step workflows.
• Build and improve automated testing infrastructure for statistically grounded analysis.
• Test novel abuse risks in agentic systems, such as indirect prompt injection and agent hijacking.
• Translate security findings into clear risk assessments and actionable recommendations for cross-functional partners.

Culture

Requirements