Principal AI Creative Director, Pod

Amazon Amazon · Big Tech · NY +1 · Design

This role is for a Principal AI Creative Director focused on building a generative AI system for Amazon Ads. The core responsibility is to act as the 'human judgment layer', defining and decomposing creative direction, brand intelligence, and quality standards into machine-readable primitives. This involves leading the systemization of taste, defining aesthetic moderation frameworks, and building prototypes to shape AI-powered creative experiences for advertisers. The role requires deep fluency in AI creation tools and the ability to translate creative vision into system requirements, influencing design, engineering, and product leadership.

What you'd actually do

  1. Define and own the creative quality bar for AI-generated assets across POD's full portfolio, including images, video, copy, and multi-format compositions.
  2. Lead the Systemization of Taste: Break down what makes great creative into its constituent elements: narrative structure, color theory, compositional hierarchy, emotional arc, pacing, and brand signal.
  3. Produce high-quality creative content independently across modalities, including image, video, and copy, using the leading AI creation tools.
  4. Partner with design, product, and engineering to define what new AI-powered creative experiences should look and feel like for advertisers.
  5. Build and own the aesthetic moderation framework for POD: the multi-dimensional quality scorecard that defines what "premium creative" means in measurable, auditable terms across visual quality, brand alignment, emotional resonance, and format appropriateness.

Skills

Required

  • Creative direction experience
  • Portfolio demonstrating exceptional craft across visual design, brand identity, and campaign development
  • Experience defining and implementing creative quality standards
  • Deep understanding of generative AI capabilities and limitations
  • Fluency in using leading AI creation tools (text-to-image, text-to-video, LLMs)
  • Ability to decompose complex creative concepts into teachable, machine-readable primitives
  • Experience in building and owning quality scorecards or moderation frameworks
  • Strong communication and cross-functional leadership skills
  • Experience prototyping AI-driven creative experiences

Nice to have

  • Experience in advertising technology
  • Familiarity with DCO (Dynamic Creative Optimization)
  • Understanding of prompt engineering and LLM-based agentic workflows
  • Experience with data annotation schemas and model training inputs

What the JD emphasized

  • creative direction
  • brand intelligence
  • systematic thinking
  • decomposing that understanding into primitives
  • primitives are what our systems learn from, replicate, and eventually govern at scale
  • aesthetic moderation
  • quality bar
  • brand-safe at scale
  • senior creative editor
  • decomposable, teachable terms
  • structured frameworks
  • prompt libraries
  • creative briefs
  • quality rubrics
  • annotation schemas
  • model training
  • DCO reward signals
  • creative tooling
  • machine-readable inputs
  • working creative prototypes
  • demo-driven culture
  • prototypes move decisions
  • text-to-image
  • text-to-video
  • LLM-powered copy and agentic workflows
  • AI-powered creative experiences
  • experience design for advertiser-facing tools
  • Creative Studio
  • Video Generator
  • creative agent
  • legible, trustworthy, and creatively empowering
  • aesthetic moderation framework
  • multi-dimensional quality scorecard
  • measurable, auditable terms
  • governance processes
  • quality thresholds
  • brand safety standards
  • editorial review workflows
  • scale without requiring human review of every asset
  • brand requirements into creative direction
  • AI systems can operationalize and enforce
  • creative authority
  • cross-functional conversations
  • demonstrated quality and creative vision
  • high creative bar
  • personalized creative variants
  • quality and brand coherence at impression level

Other signals

  • Generative AI is the unlock
  • personalized, brand-aligned creative a default capability
  • define what great looks like inside it and to make that definition teachable, scalable, and machine-readable
  • human judgment layer at the center of Amazon's AI creative system
  • creative direction: the taste, the standards, the brand intelligence, and the systematic thinking that ensures our system produces creative that builds brands, not just optimizes clicks
  • decomposing that understanding into primitives: the story arc, the color logic, the compositional tension, the emotional beat, the brand signal
  • primitives are what our systems learn from, replicate, and eventually govern at scale
  • Define the primitives that power aesthetic moderation: the signals and thresholds that allow our systems to evaluate, filter, and rank AI-generated content without requiring human review of every asset
  • build working creative prototypes that demonstrate new customer experiences, surface trade-offs, and create decision points for design, product, and engineering leadership
  • Define what new AI-powered creative experiences should look and feel like for advertisers
  • How should a brand interact with a generative creative system? What does a great AI-assisted creative workflow feel like?
  • Build and own the aesthetic moderation framework for POD: the multi-dimensional quality scorecard that defines what "premium creative" means in measurable, auditable terms across visual quality, brand alignment, emotional resonance, and format appropriateness
  • Establish governance processes, including quality thresholds, brand safety standards, and editorial review workflows, that scale without requiring human review of every asset.