Applied ML Scientist - Security Risk & Graph Analytics

Amazon Science; New York (NY); On-site; USD 184,000 - 249,000; Full time

Job summary

Amazon Science’s DNA team seeks an Applied Scientist III to invent and build ML-driven systems that prioritize and mitigate application security risk at scale. You will work with large datasets spanning app metadata, code signals, and vulnerability findings to develop impactful models.

You will lead production ML pipelines on AWS, extend graph-based models, and collaborate with security engineers, data engineers, and TPMs to convert research into actionable intelligence for security reviews.

Qualifications

• 3+ years building ML models for business applications.
• PhD or master's with 6+ years of applied research.
• Experience programming in Java, C++, Python.
• Experience with neural deep learning methods.

Responsibilities

• Lead design, development, and deployment of ML models for security prioritization and risk assessment.
• Frame ambiguous problems into well-defined scientific challenges and propose novel approaches.
• Architect production-grade ML pipelines on AWS services (S3, Glue, SageMaker).
• Develop graph-based models capturing relationships between apps, services, teams and findings.
• Drive scientific agenda with experiments and rigorous evaluation.
• Translate model outputs into actionable security insights with engineers and TPMs.
• Establish best practices for reproducibility and publish results when appropriate.

Skills

ML models

Java

Python

C++

Deep learning

Distributed systems

Education

PhD or Master’s + 6+ years applied research

Tools

TensorFlow

scikit-learn

Spark MLlib

MxNet

NumPy

SciPy

Job description

Amazon Science’s DNA team seeks an Applied Scientist III to invent and build ML-driven systems that prioritize and mitigate application security risk at scale. You will work with large datasets spanning app metadata, code signals, and vulnerability findings to develop impactful models.

You will lead production ML pipelines on AWS, extend graph-based models, and collaborate with security engineers, data engineers, and TPMs to convert research into actionable intelligence for security reviews.