Security & Test Engineer

You will serve as the continuous adversarial tester for a critical AI system, playing an active role throughout the entire 9-week development cycle. Rather than performing a final security review at the end, you will test every safety-critical component as it is built. Your written approval is essential and acts as the gatekeeper before the AI agent can interact with real financial data.

Key Responsibilities:
- Design and execute adversarial test scenarios on each safety-critical component as it is developed, ensuring testing is integrated continuously rather than deferred until project completion.
- Attempt to exploit self-approval vulnerabilities within the authorization service and approval backend.
- Test for access-control bypass vulnerabilities in the document retrieval system.
- Conduct duplicate-submission and stale-session attack attempts against the SAP connector and tool execution gateway.
- Explore bounds-bypass exploits targeting the agent configuration and template system.
- Provide a clear, written, and dated sign-off that must be approved before any real (non-sandbox) financial write actions are permitted to go live.

Required Qualifications:
- Minimum of 5 years’ experience in distributed systems reliability, including knowledge of circuit breakers and idempotent retries, combined with strong instincts for security testing.
- Ability to work closely and continuously with a fast-paced engineering team throughout the development lifecycle, rather than solely at project completion.
- Excellent written communication skills, with the ability to produce clear, defensible sign-off documentation for security decisions.

Preferred Qualifications:
- Experience in adversarial or red-team testing of systems that control access to real financial or otherwise consequential actions.
- Specific expertise in testing authorization, approval workflows, or audit-logging systems.

Work Location:
This role requires in-person attendance.

This position offers a unique opportunity to be deeply embedded in the development process, ensuring the highest security standards are met before deployment. Your work will directly impact the safety and integrity of financial data interactions within a cutting-edge AI environment.