Vals AI
Member of Technical Staff - Research, Cyber
$130K–$185K
San Francisco · On-site · Staff · Full-time · Posted Oct 21, 2026
About the job
Designs and researches cybersecurity benchmarks to evaluate real-world capabilities and risks of large language models (LLMs).
What you'll do
- Design, develop, and validate cybersecurity benchmarks for challenging real-world capabilities.
- Analyze model performance, communicate results, and publish research findings.
- Collaborate with model labs, enterprise security teams, and infrastructure teams to implement benchmarks at scale.
What they ask for
- Hands-on cybersecurity experience, such as security engineering, penetration testing, red teaming, or vulnerability research.
- Strong grounding in operating systems, cloud infrastructure, and networking.
- Interest in rigorous benchmarking and experience using LLMs in security work.
- Ability to communicate research clearly and collaborate in research teams.
Nice to have
- Prior work on a security team at a security company or in-house, and/or on an AI or ML team at a security company.
- Security-community involvement, such as CTF teams, conference talks, open-source tools, research, papers, or technical write-ups.
Pay
$130K – $185K • Offers Equity
Tools
LLMs, Python, Django, React, AWS, CDK
AI focus
Develop and evaluate benchmarks measuring LLM cybersecurity capabilities and risks.
Model-written summary · the original posting is authoritative