5 months ago
Tel Aviv-Yafo, IsraelMid Level
Responsibilities
- Build infrastructure and experiments to evaluate model capabilities and mitigate risks in frontier AI systems.
- Develop evaluation pipelines, security-focused testing frameworks, controlled environments, and agent frameworks.
- Design APIs, analysis and visualization tools, model-evasion challenges, and task-completion agents.
- Publish research and/or deliver research findings to customers.
Requirements
- Strong production programming skills and experience.
- Strong problem-solving and analytical skills.
- Ability to work effectively in a multidisciplinary team and adapt to rapidly evolving challenges.
- Interest in AI and cybersecurity; machine learning or cybersecurity experience is a plus.
- Concern for the societal impacts of the work.
Categories
About Irregular
Irregular builds security testing platforms and services to find and fix vulnerabilities in advanced AI systems before deployment, serving AI labs and enterprises adopting AI. Its high-fidelity research environments simulate real-world attack scenarios and inform defenses, monitoring, and SecOps controls across cloud and engineering systems. Founded in 2023 and privately held, Irregular was previously known as Pattern Labs.