SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Netskope is seeking a Senior AI Quality & Red Team Engineer to lead the testing and validation of AI agents before production deployment. In this role, you will own the automated evaluation suite that every agent must pass, designing and implementing adversarial test scenarios including prompt injection attempts, sycophancy checks, and multi-turn attacks. You will build repeatable, scalable processes to stress-test agents at fleet scale rather than manual one-off exercises.
Key responsibilities include: designing and automating adversarial test scenarios that run on every relevant code change; owning the "break it on purpose" phase for new agents by attempting data extraction, boundary violations, and synthetic probe exploitation; partnering with data stewards to ensure test scenarios reflect actual data sensitivity and risk; establishing clear pass/fail criteria for each agent capability and automating evaluation thresholds; maintaining a guardrail and negative-test catalog across the platform; producing audit-ready evidence automatically as part of the CI/CD pipeline; and tracking fleet-wide drift over time to catch slow-moving problems before they compound.
You will work in a cloud security context where AI agents interact with sensitive data and systems. The role requires building infrastructure that scales from dozens to hundreds of agents, with emphasis on automation, repeatability, and compliance-ready documentation. This is not a role that babysits individual test runs—you are architecting the harness that keeps the entire AI fleet secure and resilient.
