Staff Product Security Engineer | Okta

What you'll do

• Conduct offensive security research on agentic AI systems to identify risks like prompt injection and privilege escalation.
• Perform security assessments of Okta's AI platforms across design, code, and runtime phases.
• Build reusable security tooling that enhances the Product Security team’s capabilities.
• Evaluate AI security tools through benchmarking harnesses designed for assessing vendor solutions.
• Develop threat models for agentic architectures and LLM-integrated services to guide engineering teams.

What we're looking for

• 7+ years of experience in information security with depth in application security and offensive research.
• Hands-on experience assessing LLM-integrated systems and agentic AI architectures, identifying real vulnerabilities.
• Strong offensive mindset to model adversary behavior and construct risk scenarios for agentic systems.
• Experience building and using security tooling and automation scripts for other engineers.
• Proficiency in Python and one additional language (Go, Java, TypeScript, C/C++) for security assessments.
• Advanced threat modeling, manual code review, and penetration testing skills for complex distributed systems.