Senior Research Engineer - AI Security
Core
Design, deploy, and operate production-scale ML systems to protect Copilot experiences from threats like prompt injections and adversarial inputs.
Role type
Senior Research Engineer - AI Security
Builds
Adaptive detection and policy models, evaluation frameworks, and secure orchestration frameworks for agent delegation and tool calling.
Domain
AI Security / Large Language Model Safety
Deliverable
production ML models
Required skills
ML fundamentals (classification, anomaly detection, representation learning), end-to-end ML pipeline design, adversarial testing, telemetry/observability pipelines, agentic AI risk understanding, distributed systems, secure system design
Preferred skills
Experience with C/C++/C#/Java/JavaScript/Python, AI safety/trust/security adjacent problems, clear communication of complex concepts
Technologies
C, C++, C#, Java, JavaScript, Python
Responsibilities
Develop adaptive detection models learning from evolving attacker behavior, build evaluation frameworks for adversarial testing and red-teaming, partner with security teams to integrate ML defenses, monitor telemetry to reduce false positives/negatives, conduct rigorous experimentation to quantify security effectiveness
Seniority
Senior, hands-on IC