Senior Applied Scientist, AWS Security

We build AI-powered tooling that enables security operations to scale with AWS's growth. Our portfolio includes generative AI incident response assistants, natural language‑driven response, detection enrichment pipelines, and security data analytics platforms. Security analysts depend on these systems around the clock.

We are hiring a Senior Applied Scientist to own the science strategy for our AI security response platform. You will define and execute the machine learning and AI roadmap across our service portfolio, from large language model‑powered incident triage to anomaly detection in security telemetry. You will extend and invent techniques at the product level, partnering with software and security engineers to bring models from research into production systems that operate 24/7/365.

You will be the scientific authority on the team, expected to teach, mentor, and set the technical bar for how we apply AI to security operations problems.

This role requires deep expertise in natural language processing, generative AI, or a closely related discipline, combined with a demonstrated ability to translate scientific methods into production systems that solve real business problems. You will operate in high‑ambiguity, high‑consequence domains where your scientific judgment directly affects security outcomes for AWS.

• Define and own the science strategy for the team's AI‑powered security automation portfolio, including model selection, evaluation methodology, and research direction.
• Design and implement LLM‑powered systems for security incident triage, including retrieval‑augmented generation, prompt engineering, and fine‑tuning approaches that improve recommendation accuracy and reduce analyst toil.
• Build anomaly detection and classification models across security telemetry data sources to surface threats, reduce false positives, and prioritize analyst attention.
• Partner with software engineers to move models from experimentation to production. Define system‑level technical requirements, guide adaptation to meet production constraints, and own model performance in deployment.
• Develop evaluation frameworks and metrics that measure model effectiveness against security outcomes, not just standard ML benchmarks.
• Mentor software and security engineers on ML best practices and raise the science bar across the team through design reviews, code reviews, and knowledge sharing.

You start by reviewing model performance dashboards for overnight incident triage recommendations, investigating a drift in precision for a specific detection category. Mid‑morning, you lead a design review for a retrieval‑augmented generation pipeline that will surface relevant runbooks during security incidents. After lunch, you pair with a security analyst to label edge cases that your current model misclassifies, turning operational feedback into training signal.

You close the day writing an experiment plan to evaluate a new embedding approach for security log similarity, then sync with your manager on the quarterly science roadmap.

This team operates within AWS Security in a 24/7/365 organization that protects AWS's global cloud infrastructure. The team builds AI‑powered security automation, data analytics platforms, and incident response tooling that security analysts depend on around the clock. We work at the intersection of machine learning, generative AI, and security operations. Our mission: give every security analyst the intelligent tooling they need to stay ahead of threats at AWS scale.

• 3+ years of building machine learning models for business application experience
• PhD, or Master's degree and 6+ years of applied research experience
• Experience programming in Java, C++, Python or related language
• Experience with neural deep learning methods and machine learning

• Experience with modeling tools such as R, scikit‑learn, Spark MLLib, MxNet, Tensorflow, numpy, scipy etc.
• Experience with large‑scale machine learning systems such as profiling and debugging and understanding of system performance and scalability
• Experience in applied research

A…