Applied Scientist: Identity Security & Abuse Prevention

Amazon Science; Seattle (WA); On-site; USD 142,000 - 193,200; Full time

Job summary

Amazon Science's Identity Security & Abuse Prevention (ISAP) team is seeking an Applied Scientist to build, deploy, and own machine learning systems that detect abuse and automate enforcement across Amazon’s identity landscape.

You will design and deploy production ML models, advance detection for emerging threats, and mentor junior scientists while partnering with security engineers and data engineers to implement end-to-end detection and enforcement pipelines using GenAI and LLM-based

Qualifications

• PhD or Master’s degree with 4+ years CS/ML experience
• Experience publishing in top conferences/journals or patents
• Proficiency in Java, Python, or C++ with production ML systems
• Strong background in threat detection, graph analytics, and anomaly detection

Responsibilities

• Design, develop, and deploy production ML systems for abuse pattern detection, anomaly detection, threat classification, and automated enforcement across multiple Amazon verticals
• Independently frame security problems into well-defined scientific questions and drive from hypothesis to production
• Own and improve detection models end-to-end: monitor drift, diagnose degradation, retrain, extend coverage
• Build and maintain graph-based entity analysis and identity resolution systems
• Design experiments (A/B tests, offline evaluation) to measure model performance and business impact
• Architect GenAI and LLM-based solutions for investigation automation and knowledge retrieval
• Contribute to scientific roadmap by proposing high-value detection opportunities
• Publish research findings in internal/external venues and collaborate with security/data engineers

Skills

Java

C++

Python

Graph neural networks

Anomaly detection

Education

PhD in CS/CE/ML

MS in CS/CE/ML

Tools

Hadoop

Spark

LLMs

RAG systems

Job description

Amazon Science's Identity Security & Abuse Prevention (ISAP) team is seeking an Applied Scientist to build, deploy, and own machine learning systems that detect abuse and automate enforcement across Amazon’s identity landscape.

You will design and deploy production ML models, advance detection for emerging threats, and mentor junior scientists while partnering with security engineers and data engineers to implement end-to-end detection and enforcement pipelines using GenAI and LLM-based