Senior Applied Scientist, Leo Security

About the role

Applied research and production role for security analytics across Amazon Leo satellite infrastructure, combining graph-based neurosymbolic reasoning, LLM agents, RL fine-tuning, behavioral anomaly detection, and evaluation. Export-control eligibility is limited to U.S. citizens/nationals, permanent residents, refugees, or asylees.

Requirements

• 3+ years building machine learning models for business applications
• PhD or master degree plus 6+ years applied research experience
• Programming in Java, C++, Python, or related language
• Neural deep learning and machine learning
• Graph algorithms, knowledge graphs, unsupervised learning, NLP, and embedding retrieval
• Evaluation frameworks for AI systems with human-judgment ground truth
• Export-control eligibility limited to U.S. citizens/nationals, permanent residents, refugees, or asylees

What you'll do

• Build production neurosymbolic systems combining graph reasoning and LLM agents
• Develop RL and fine-tuning pipelines for security models
• Detect threat actor behavior and evaluate models before production
• Build multi-agent systems for security triage and response
• Own research through deployment and define impact metrics

Where this was posted

• LinkedIn

  last seen 2h ago
  View on LinkedIn