Applied Scientist Ii, Visual Search Science

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

Applied Scientist II role focused on building generative AI and multimodal search systems for Amazon's visual search experience. The role involves designing, training, and optimizing text-to-image generation models, developing multimodal retrieval systems, building LLM-based classifiers for intent and safety, and architecting GPU-intensive inference pipelines. It operates at Amazon scale, serving hundreds of millions of customers.

What you'd actually do

  1. design, train, and optimize generative AI models for real-time product image generation, ensuring outputs meet strict latency requirements while maintaining high visual quality and query alignment.
  2. develop multimodal retrieval systems that connect AI-generated images to Amazon's billions-scale product catalog, optimizing for recall and ranking relevance across product categories.
  3. building LLM-based classifiers for visual intent detection, query understanding, and safety filtering within real-time latency budgets.
  4. advance AI safety science through defense-in-depth approaches including embedding-space classifiers, adversarial data engines, and post-generation content moderation.
  5. design and execute large-scale online experiments to measure impact on customer engagement, search success, and business metrics, defining evaluation frameworks that combine automated metrics with human judgment.

Skills

Required

  • PhD, or Master's degree and 4+ years of CS, CE, ML or related field experience
  • Experience programming in Java, C++, Python or related language
  • Experience in state-of-the-art deep learning models architecture design and deep learning training and optimization and model pruning

Nice to have

  • Experience using Unix/Linux
  • Experience in professional software development
  • Experience developing and implementing deep learning algorithms, particularly with respect to computer vision algorithms
  • Have publications on top-tier conferences, such as CVPR, ICCV, ECCV or NeurIPS

What the JD emphasized

  • strict latency requirements
  • billions-scale product catalog
  • real-time latency budgets
  • AI safety science
  • large-scale online experiments
  • GPU-intensive inference pipelines

Other signals

  • generative AI
  • multimodal understanding
  • diffusion models
  • LLMs
  • billions-scale catalog
  • real-time generation
  • real-time latency
  • AI safety
  • large-scale online experiments
  • GPU-intensive inference pipelines