AI Security Researcher

What you'll do

• Design red-team agents, their tools and MCP servers, and repeatable offensive skills; review agent runs and address capability gaps.
• Extend agents across full attack chains and Windows, Active Directory, on-prem, cloud, and network environments; build evasion and detection-validation loops.
• Operate cyber ranges, create benchmarks, verify high-severity attack paths in customer environments, and share research with the security community.

What they ask for

• 6+ years of experience in offensive security, red team operations, or adversary emulation.
• Hands-on experience running end-to-end operations across Windows, Active Directory, and on-prem environments.
• Depth in Command and Control frameworks and payload development, with real EDR detection-evasion experience.
• Code-writing ability to make tradecraft repeatable and automatable, plus experience operating realistic test environments or cyber ranges with production defenses.

Nice to have

• Cloud or network attack experience; experience building security evaluations or benchmarks.
• Open-source tooling, conference talks, or published offensive-security research; experience with AI agents, LLM tooling, or agentic workflows.

Tools

Model Context Protocol (MCP), Mythic, Windows, Active Directory, EDR, Security Information and Event Management, Ludus, Python

AI focus

Build and evaluate offensive AI agents that convert red-team tradecraft into autonomous, agent-executable capabilities and help train defensive agents.