What if your security expertise could help the next generation of AI systems understand how software actually breaks in the real world?
We're partnering with leading AI research labs to build training data that teaches AI how to think about application security — from identifying real exploitability to recommending sound remediation strategies. If you've spent time in the trenches of application, product, or software security, your instincts and knowledge are exactly what we need.
Analyze realistic application security scenarios spanning code, APIs, and system behavior
Classify vulnerabilities including authentication flaws, injection risks, access control issues, and business logic weaknesses
Evaluate secure coding practices and assess the quality of proposed remediation strategies
Distinguish genuine, exploitable risk from theoretical or overstated findings
Create and validate security-focused reasoning datasets that help AI systems accurately assess application risk
Provide structured, clearly written security analysis that can be used as high-quality training signal
Who You Are
2+ years of hands-on experience in application security, product security, or software security roles
Solid familiarity with the OWASP Top 10 and common vulnerabilities found in production systems
Strong ability to reason about real-world exploitability — not just textbook definitions
Clear, structured written communication skills when explaining security concepts and decisions
Self-motivated and comfortable working independently on asynchronous, task-based assignments
Nice to Have
Experience with security code review, penetration testing, or bug bounty programs
Familiarity with secure development frameworks or threat modeling
Background in writing security documentation, reports, or advisories