Sr. Applied Science Manager, Agentic AI Ads, Sponsored Products and Brands

Amazon Amazon · Big Tech · Palo Alto, CA · Applied Science

Lead a new applied science organization focused on building agentic AI systems for advertising campaigns. This role involves defining the scientific vision, research agenda, model architectures, and evaluation frameworks for LLM-based agents, multi-step planning, tool use, RAG, and RLHF, with a focus on delivering measurable value and transforming the advertiser journey. The role requires building and mentoring a team, partnering with cross-functional teams, and driving execution from research to production at scale.

What you'd actually do

  1. Build, mentor, and lead a new, high-performing applied science organization of applied scientists, research scientists, and engineers, fostering a culture of scientific excellence, innovation, customer obsession, and ownership.
  2. Define, own, and drive the long-term scientific and product vision and research strategy for agentic AI-powered advertising experiences across Sponsored Products and Brands—identifying the highest-impact research problems and charting a path from exploration to production.
  3. Lead the research, design, and development of novel agentic AI models and systems—including LLM-based agent architectures, multi-agent orchestration, planning and reasoning frameworks, tool-use mechanisms, and retrieval-augmented generation pipelines—that deliver measurable value for advertisers and create delightful, intuitive experiences.
  4. Establish rigorous scientific methodology and evaluation frameworks for assessing agent performance, reliability, safety, and advertiser outcomes, setting a high bar for experimentation, reproducibility, and offline-to-online consistency.
  5. Partner closely with senior business, engineering, and product leaders across Amazon Advertising to translate advertiser pain points and business opportunities into well-defined science problems, and deliver cohesive, production-ready solutions that drive advertiser success.

Skills

Required

  • Ph.D. in Computer Science, Machine Learning, Artificial Intelligence, Statistics, or a related quantitative field, or equivalent combination of M.S. and relevant experience.
  • 10+ years experience in applied science, machine learning, GenAI/LLMs or AI research, with a strong track record of developing a

Nice to have

  • LLM-based agent architectures
  • multi-step planning and tool use
  • retrieval-augmented generation
  • reinforcement learning from human and business feedback
  • robust evaluation methodologies for agentic systems
  • multi-agent orchestration
  • planning and reasoning frameworks
  • tool-use mechanisms

What the JD emphasized

  • end-to-end ownership
  • single-threaded applied science leader
  • build and guide a dedicated team
  • advancing the science behind intelligent agents
  • simplify campaign creation
  • automate optimization decisions
  • autonomous reasoning and planning
  • personalized advertising strategies at scale
  • pioneer novel approaches
  • LLM-based agent architectures
  • multi-step planning and tool use
  • retrieval-augmented generation
  • reinforcement learning from human and business feedback
  • robust evaluation methodologies for agentic systems
  • proactively identify and tackle the next generation of AI-powered advertising experiences
  • democratize advertising success
  • Ph.D. in Computer Science, Machine Learning, Artificial Intelligence, Statistics, or a related quantitative field, or equivalent combination of M.S. and relevant experience.
  • 10+ years experience in applied science, machine learning, GenAI/LLMs or AI research, with a strong track record of developing a

Other signals

  • Pioneering breakthrough agentic AI systems
  • Defining the scientific charter for transformative impact
  • Building and guiding a dedicated team of applied scientists, research scientists, and machine learning engineers
  • Advancing the science behind intelligent agents that simplify campaign creation, automate optimization decisions through autonomous reasoning and planning, and deliver personalized advertising strategies at scale
  • Pioneer novel approaches in areas such as LLM-based agent architectures, multi-step planning and tool use, retrieval-augmented generation, reinforcement learning from human and business feedback, and robust evaluation methodologies for agentic systems