Principal Applied Scientist, Trusted Supply, Amazon Ads

Company: Amazon
Apply for the Principal Applied Scientist, Trusted Supply, Amazon Ads
Location: London
Job Description:

Overview

As Principal Applied Scientist, you will define the science vision for Brand Safety, Suitability, and Risk Hunting to safeguard advertiser trust. You’ll lead cross-functional programs and shape scalable ML solutions processing billions of impressions daily, driving advertiser confidence and product impact. You’ll mentor a high-performing team, influence industry practices, and translate science into advertiser-facing capabilities that improve transparency and quality.

Responsibilities

  • Set multi-year science directions and publication roadmap for Brand Safety, MFA detection, traffic quality, and viewability
  • Lead cross-team collaboration between science and engineering to ship low-latency, scalable models
  • Act as a thought leader on industry shifts (privacy, adversaries, GenAI threats) and represent Amazon in forums
  • Hire, mentor, and grow a high-performing applied science team with a culture of rigor and experimentation
  • Partner with engineering leadership to deploy production-grade systems handling billions of requests per day
  • Translate scientific capabilities into product improvements (controls, reports, quality guarantees) and quantify business impact

Key requirements

  • Ph.D. in Computer Science, Machine Learning, Statistics, or a highly quantitative field
  • Experience applying ML to real-world problems at scale in a science leadership role
  • Proven track record of leading and growing teams of scientists (5+ individuals)
  • Deep expertise in NLP, Computer Vision, or multi-modal learning with production impact
  • Top-tier ML/AI publication record (NeurIPS, ICML, KDD, WWW, ACL, EMNLP, CVPR, etc.)
  • Experience with large-scale distributed ML systems processing terabytes of data
  • Expert Python and at least one systems language (Java, C++, Scala)
  • leadership and mentorship
  • strategic thinking
  • cross-functional collaboration
  • NLP
  • Computer Vision
  • multimodal learning

…

Posted: September 30th, 2026