Sr Manager, Applied Science, Creative Intelligence

Amazon Amazon · Big Tech · NY +1 · Applied Science

This role leads a science organization focused on Dynamic Creative Optimization (DCO) and Creative Brain (CB) for Amazon Ads. The core responsibility is to own and drive the models, algorithms, and research that personalize ad creative in real-time to improve advertiser performance, focusing on conversion and long-term value. The role involves managing a team of applied scientists and data scientists to build a closed-loop system that continuously learns from each impression to improve creative decisions. Key objectives include compressing learning loops, expanding optimization coverage, solving cold-start problems, building persistent memory layers, developing causal inference for creative components, creating cross-advertiser priors, designing representation architectures for creative quality reasoning, and owning quality science (defect detection, compliance, aesthetics). The role also involves defining science strategy, leading competitive analysis, ensuring rapid translation of research to production, and managing a team of 8+ scientists.

What you'd actually do

  1. Own the models that personalize ad creative at serving time across all formats.
  2. Drive optimization beyond clicks toward conversion, consideration, and long-term advertiser value.
  3. Compress the learning loop from days to hours.
  4. Expand model-driven optimization from partial to full coverage.
  5. Solve cold-start and self-competition problems.

Skills

Required

  • PhD in Machine Learning, Statistics, or Computer Science (or MS + 10 years equivalent applied experience)
  • 8+ years building and shipping production ML systems
  • 5+ years managing teams of 10+ scientists or ML engineers
  • Experience with real-time serving systems and online learning at scale
  • Track record of measurable business impact from deployed models

Nice to have

  • Experience in recommendation systems, causal inference, or multi-armed bandits
  • Background in ad-tech creative optimization or dynamic content personalization
  • Prior role at a peer platform (Meta, Google, TikTok) in ads or creative science
  • Strong point of view on where creative optimization intersects with auction design and platform economics
  • Comfort defining strategy in ambiguity rather than executing a handed roadmap
  • Publications or patents in relevant areas

What the JD emphasized

  • building and shipping production ML systems
  • managing teams of 10+ scientists or ML engineers
  • real-time serving systems and online learning at scale
  • measurable business impact from deployed models

Other signals

  • personalization at impression level
  • measurable advertiser performance lift
  • closed-loop system
  • real-time serving systems and online learning at scale