Remote Senior AI Security Engineer — Attack Simulation

Australia-Employment; Los Angeles (CA); Remote; USD 250,000 - 400,000; Full time

Job summary

Unknown AI Cybersecurity Firm is seeking a Lead Engineer to build an AI Attack Simulation platform in a remote US setting, reporting directly to a small autonomous pod. You will drive architecture, implement offensive security techniques, and ship resilient production software across AI models and agentic workflows.

Ideal candidates combine hands-on software engineering with attacker mindset, delivering scalable security testing for AI environments and LLM-powered systems while collaborating

Qualifications

• Experience in production software engineering and building scalable systems.
• Strong background in offensive security, attack simulation, or red-teaming.
• Familiarity with AI/ML security concepts and production workflows.

Responsibilities

• Take ownership of an AI Attack Simulation platform from design to production.
• Turn attack methods and security testing techniques into scalable software.
• Collaborate with product, customers, and engineers to validate security solutions.
• Prototype, validate, and iterate quickly in a fast-moving startup environment.
• Build and maintain production-grade systems that measure security outcomes.
• Engage with customers to ensure solutions meet real-world needs.

Skills

Offensive Security

Production Software Engineering

AI/ML Security

Agentic Systems

Tools

Go

Python

Redis

OpenSearch

Kubernetes

AWS

Microservices

AI/LLM Systems

Agentic AI

Job description

Unknown AI Cybersecurity Firm is seeking a Lead Engineer to build an AI Attack Simulation platform in a remote US setting, reporting directly to a small autonomous pod. You will drive architecture, implement offensive security techniques, and ship resilient production software across AI models and agentic workflows.

Ideal candidates combine hands-on software engineering with attacker mindset, delivering scalable security testing for AI environments and LLM-powered systems while collaborating