Security Software Engineer

The role

You will research and develop offensive agentic capabilities for Sybil. You will hunt for vulnerabilities and build end-to-end customer-facing features using LLM-powered tooling.

What you will do

• Research and develop offensive agentic capabilities that become core Sybil features
• Hunt for real vulnerabilities across a range of target environments
• Translate findings into repeatable, automated techniques
• Build and ship end-to-end customer-facing features from research spike to production deployment
• Design and implement LLM-powered tooling
• Identify gaps in offensive coverage and propose technical initiatives
• Contribute to technical design discussions
• Review teammates' work
• Operate with ownership to define problems and build solutions

Required qualifications

• 4 or more years of software engineering experience, with meaningful time in offensive security contexts
• Hands-on experience finding vulnerabilities through red teaming, bug bounty, CTF, or production pentesting
• Strong Python fundamentals including designing, deploying, and maintaining backend systems in production
• Comfort with modern JavaScript and React
• Experience building with LLMs or generative AI tools, including prompt engineering and integrating model outputs into real workflows
• Familiarity with DevOps and infrastructure concepts
• Strong written communication

Benefits

• Meaningful equity

About the company

RunSybil is building Sybil, an AI-driven pentester that discovers vulnerabilities. The team is composed of experts from OpenAI, Meta, Mandiant, Palantir, Cruise, Trail of Bits, and Aptiv.