About OpenTrain
OpenTrain is the unified marketplace where AI training freelancers find and consolidate opportunities across platforms, build an AI training portfolio, and turn specialized knowledge into a durable freelance career. We help candidates discover high-quality AI training, evaluation, and data work without scattering their track record across disconnected sites.
About this role
OpenTrain is recruiting for a Principal Cybersecurity Expert to support AI training work focused on Cybersecurity AI Model Evaluation. This role uses Cybersecurity data, investigation cases, and decision guidance so AI systems can learn from structured human expertise, careful review, and clear written feedback. Prior AI training experience is not required unless stated elsewhere on this job page; the most important qualification is strong real-world knowledge of the subject area.
What you will do
- evaluate model outputs, apply project rubrics, and document evidence-based judgments
- Review source materials and model outputs against project guidelines
- Explain judgments clearly so the feedback can improve model behavior
- Maintain consistency and quality across repeated review tasks
Required skills
- 8 to 15 years of hands-on cybersecurity experience in areas such as security operations, incident response, threat intelligence, application security, penetration testing or vulnerability research.
- Experience leading or mentoring analysts or engineers, including setting standards and reviewing their work.
- Working experience with large language models: using them in security workflows, evaluating their outputs, red-teaming them or contributing to AI training data.
- Offensive and defensive range deep enough to judge others' analysis across domains.
- Sound judgment on security policy, responsible disclosure and dual-use content.
- Clear written English for writing guidelines others follow.
Helpful background
- Familiarity with how LLMs fail on security tasks.
- Certifications such as OSCP, OSCE, GXPN, GCIH or CISSP.
- Experience building playbooks, detection content or security training material.
- Prior work on AI safety, trust and safety, or data annotation programs.
Schedule and pay
Full-time availability for 6 to 12 months with no competing full-time commitments.
Compensation is handled through the OpenTrain project budget fields shown on this job page.
Why work with OpenTrain
OpenTrain gives freelancers one place to find AI training work, build a unified AI training portfolio, and use that work history to qualify for more opportunities over time. For domain experts, language specialists, writers, engineers, and reviewers, AI training is a growing market where specialized judgment can become a flexible and repeatable source of income.