Guardrail Engineering & Operations
• Implement, tune, and operate detection and policy content for enterprise AI guardrails (for example sensitive data and secrets detection, prompt-injection defenses, and tool-call policies) across AI assistants, coding agents, and internally built agents.
• Triage findings, classify true and false positives against an agreed rubric, and produce the evidence needed to move controls from monitoring to blocking with minimal user impact.
• Maintain runbooks, configuration as code, and test suites for guardrail rules and policies; validate changes before promotion to production.
• Participate in the operational cadence of the service, including regular tuning reviews, change management, and support rotations as needed.
AI Systems Assessment & Testing
• Assess AI applications, agents, integrations (including MCP servers and tool connectors), and third-party AI products to document what they can access and do, and where controls are needed.
• Execute test plans for prompt injection (direct and indirect), sensitive data exposure, excessive permissions, and unsafe agent actions; reproduce issues and write clear, actionable findings.
• Contribute to threat models and security acceptance criteria for AI systems using frameworks such as OWASP Top 10 for LLM Applications and MITRE ATLAS.
• Track remediation with owning teams and verify that fixes are effective.
Visibility, Detection & Response
• Build and maintain telemetry pipelines, analytics, and dashboards for AI usage and agent activity so teams can see where AI is used, what it can access, and whether controls are working.
• Develop detections and alerts for AI-specific events such as data exfiltration through AI tools, anomalous agent behavior, and unsafe tool use, and forward them to SIEM and SOAR platforms.
• Support security operations and insider-threat investigations involving AI tools by providing data, context, and analysis.
• Measure control effectiveness with agreed metrics (coverage, precision, time to detect, user impact) and report trends.
Platform Integration & Automation
• Integrate AI security tooling with cloud (AWS, Azure), endpoint, identity, data protection, and ticketing platforms through governed, authenticated APIs.
• Automate deployment, testing, and rollback of guardrail services and detection content using CI/CD and infrastructure-as-code; keep configurations version-controlled and auditable.
• Evaluate and pilot vendor and platform capabilities under the direction of senior engineers, documenting what they cover and where gaps remain.
• Apply secure engineering practices (least privilege, secrets management, logging) to the tooling the team builds and operates.
Collaboration & Continuous Learning
• Work directly with teams building AI applications and agents to help them apply security requirements and approved patterns.
• Contribute to documentation, standards, and reusable patterns; share findings with security operations, privacy, and platform engineering peers.
• Stay current on LLM and agent threats, vendor guardrail capabilities, and industry frameworks, and bring relevant practices back to the team.
• At the Principal Engineer (R3) level: lead small workstreams end to end, mentor less experienced engineers, and represent the team in cross-functional working sessions.