Staff AI Security Engineer: Red-Teaming & Certification

Wintermeyer Ventures; Verona (NJ); On-site; USD 230,000 - 400,000; Full time

Job summary

Wintermeyer Ventures in Verona, USA, seeks an experienced software engineer to join an early-stage AI risk team. You will lead end-to-end red-teaming and evaluation for customer AI agents, from scoping through remediation and certification.

The role emphasizes building test tooling, improving evaluation workflows, and turning frontier adversarial attacks into repeatable product capabilities. You will work with Python and TypeScript, own customer engagements, and contribute at scale in a hands-on

Qualifications

• At least 5 years of software engineering experience with end-to-end ownership of high-stakes technical work.
• Proficiency in Python and TypeScript with production deployments.
• Experience with AI agents or related evaluation frameworks and regular use of agentic AI tools.
• Startup experience, especially early-stage, is valued and experience in larger tech orgs is beneficial.
• Experience applying product management practices and working with customers and cross-functional stakeholders.

Responsibilities

• Refine the core engine that runs red-teaming evaluations.
• Own technical scoping, reconnaissance, test execution, remediation, and certification for customer AI agents.
• Turn frontier adversarial attacks into repeatable product capabilities.
• Review and calibrate evaluations to maintain a high quality bar.
• Use agentic tools to productize and improve your workflow.
• Develop evaluation approaches for new agent capabilities, including computer use.

Skills

Python

TypeScript

End-to-end ownership

AI agents

Customer engagement

Education

Bachelor's degree in Computer Science or equivalent

Tools

CI/CD

Git

Testing frameworks

Job description

Wintermeyer Ventures in Verona, USA, seeks an experienced software engineer to join an early-stage AI risk team. You will lead end-to-end red-teaming and evaluation for customer AI agents, from scoping through remediation and certification.

The role emphasizes building test tooling, improving evaluation workflows, and turning frontier adversarial attacks into repeatable product capabilities. You will work with Python and TypeScript, own customer engagements, and contribute at scale in a hands-on