Sr. Applied Science Manager, Perfect Order Experience (poe) AI

Amazon Amazon · Big Tech · Seattle, WA · Applied Science

Senior Applied Science Manager leading a team to develop a domain-specific LLM, including pre-training, fine-tuning, and reinforcement learning. The role also involves architecting risk detection systems using multi-modal signals and influencing ranker models for product visibility. The focus is on building and scaling AI solutions for Amazon's Perfect Order Experience.

What you'd actually do

  1. Drive AI strategy and lead a team of applied scientists in developing ML solutions.
  2. Lead the end-to-end development of a domain specific LLM.
  3. Drive the development of large-scale pre-training and post-training strategies for the LLM using domain-specific datasets.
  4. Architect automated risk detection and treatment systems that combine multi-modal signals to identify product quality issues and implement optimization-based mitigation strategies.
  5. Collaborate with other science teams to develop/ influence ranker models that optimize product visibility.

Skills

Required

  • Ph.D. in Computer Science, Machine Learning, or related technical field, or equivalent practical experience
  • Experience leading and managing teams of scientists/engineers in delivering ML solutions at scale
  • Strong track record in developing and deploying production ML systems
  • Strong publication record or proven industrial innovations (e.g., patents) in ML/AI

Nice to have

  • Strong communication skills with ability to translate complex technical concepts to various audiences
  • Experience with LLM development, including pre-training, fine-tuning, and reinforcement learning
  • Knowledge of search, ranking, or recommendation systems
  • Experience with multi-modal ML systems combining text, image, and structured data

What the JD emphasized

  • Experience leading and managing teams of scientists/engineers in delivering ML solutions at scale
  • Strong track record in developing and deploying production ML systems
  • Strong publication record or proven industrial innovations (e.g., patents) in ML/AI
  • Experience with LLM development, including pre-training, fine-tuning, and reinforcement learning

Other signals

  • leading AI strategy
  • end-to-end LLM development
  • pre-training and post-training strategies
  • risk detection and treatment systems
  • multi-modal signals