Guardrails Engineer

Guardrails Engineer will build and automate guardrail testing, adversarial testing, policy validation, security checks, and evidence generation for AI agents and LLM-powered solutions. This role is responsible for helping ensure safe, compliant, and controlled deployment of Agentic AI use cases.

Key Responsibilities
Develop automated guardrail testing frameworks for LLM-powered applications, agents, tools, and workflows.
Perform adversarial testing, prompt injection testing, jailbreak testing, red-team-style testing, and policy control validation.
Validate content safety, data protection, security, responsible AI, and model risk controls.
Build repeatable evidence packages to support approval gates and governance reviews.
Collaborate with Product Owners, risk, security, observability, and engineering teams to define control requirements.
Integrate guardrail tests into platform lifecycle, CI/CD, release readiness, and production enablement processes.
Document findings, control gaps, remediation recommendations, and readiness evidence.

Required Qualifications
7+ years of engineering, security testing, QA automation, risk technology, AI safety, or platform testing experience.
5+ years of strong hands-on Python and test automation experience.
5+ years of Understanding of LLM risks, prompt injection, jailbreak attempts, content safety, data leakage, and Responsible AI controls.
5+ years of Experience creating test harnesses, validation processes, and repeatable evidence artifacts.
Ability to work with engineering, security, risk, compliance, and product stakeholders.

Required Skills / Knowledge
Python, adversarial testing, red-team testing, prompt security, responsible AI, guardrails, security testing, test harnesses, and evidence automation.
Understanding of AI governance, model risk, content policy controls, approval gates, and regulated delivery.
Ability to automate repeatable quality and safety checks for enterprise AI solutions.

Preferred Qualifications
Experience with GenAI security, AI red teaming, model risk controls, or regulated AI delivery.
Experience with CI/CD, observability, Kubernetes, cloud platforms, and enterprise security frameworks.
Banking or financial services experience.
Expected Outcomes
Automated guardrail and adversarial testing framework.
Repeatable governance evidence for AI approval gates.
Improved safety, control validation, and release confidence for agents.