Senior Staff ML Engineer, AI Safeguards - Moveworks

ServiceNow • Mountain View, CA

19h

About The Position

The Senior Staff ML Engineer will design, build, and ship safeguards for Moveworks’ agent harness. The safeguards will address prompt injection, single- and multi-turn jailbreaks, toxicity and sensitive topics, and sensitive data exposure. The role includes developing evaluation systems to measure safeguard efficacy across false and true positives using customer reports, adversarial testing, and production traffic. The engineer will collaborate with Security to turn adversarial findings into evaluation cases and fixes, and to develop mitigations that remain effective when detection misses.

The role will select, integrate, and fine-tune detection models for a stack shared by commercial and government environments. It will balance quality, latency, and cost for controls running in the request path, create configurable safeguard policies that let customers choose how sensitive topics are handled, and monitor production behavior to identify and fix regressions.

Requirements

• Typically 8 or more years of experience building production software.
• At least 3 years of applied machine learning experience, including training, evaluating, and deploying models in production systems.
• Hands-on experience developing LLM safeguards, such as guardrail models, content classifiers, prompt-injection or jailbreak detection, or safety evaluations.
• Strong model-evaluation fundamentals for agent trajectories and text classification, with attention to training- and evaluation-data quality.
• Working knowledge of how LLM safeguards are bypassed, including indirect prompt injection through tools and documents and multi-turn jailbreaks.
• Excellent Python experience.

Responsibilities

• Design, build, and ship agent-harness safeguards for prompt injection, single- and multi-turn jailbreaks, toxicity and sensitive topics, and sensitive data exposure.
• Build evaluations of safeguard efficacy, measuring false and true positives with datasets from customer reports, adversarial testing, and production traffic.
• Collaborate with the Security team to turn adversarial findings into evaluation cases and fixes, and build mitigations that remain effective when detection misses.
• Select, integrate, and fine-tune detection models for a stack shared by commercial and government environments.
• Balance quality, latency, and cost for controls that run in the request path.
• Build configurable safeguard policies that allow customers to determine how sensitive topics are handled.
• Monitor safeguard behavior in production and fix regressions.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: No Education Listed