Applied Science Manager, Advertising Trust
Search Jobs
Search by job, company or skills
Search by job, company or skills
Amazon Ads delivers advertising experiences across Amazon's owned-and-operated properties and third-party networks, reaching hundreds of millions of customers worldwide. Within Amazon Ads, Advertising Trust is the science-first organization responsible for ensuring every ad shown to customers meets Amazon's content policies - at massive scale, across all ad formats and global marketplaces.
The Ads Trust Science team builds the ML systems that automate content moderation decisions: multimodal classification, retrieval-based labeling, LLM reasoning, and agentic self-improvement architectures. This requires inventing new approaches at the intersection of computer vision, NLP, information retrieval, and generative AI.
We are seeking an Applied Science Manager to lead a team of applied scientists building next-generation content moderation intelligence. You will own the science roadmap for one of the highest-impact automation programs in Amazon Advertising, defining how multimodal content understanding, retrieval-first classification, and LLM-based reasoning combine into a production system that serves global advertising at scale.
Key job responsibilities
. Lead a team of applied scientists working across multimodal ML (vision-language models, video understanding), large-scale retrieval systems (embedding-based similarity and deduplication), and generative AI (LLM-based policy reasoning, knowledge distillation, agentic architectures, reinforcement learning).
. Define the science strategy for ads trust.
. Own end-to-end delivery of ML solutions: problem formulation, offline experimentation, online A/B testing, and production deployment. Your models directly move automation and defect metrics reported to senior leadership.
. Build and grow scientists - hire, mentor, and develop team members. Raise the science bar through structured review processes and a publication culture within Amazon.
. Partner with engineering, product, and operations teams to translate science investments into measurable automation improvements. Influence roadmaps across dependent teams.
. Communicate science strategy and results to senior leadership through narratives, technical deep-dives, and roadmap documents.
- 8+ years of applied research experience
- 4+ years of scientists or machine learning engineers management experience
- 4+ yrs in managing team of 5-15 members
- Experience building production ML systems at Internet scale, especially involving multimodal deep learning, generative AI, or large-scale retrieval
- Track record of delivering automation or classification systems with measurable business impact
- Experience with content moderation, trust & safety, or policy enforcement systems
- Publications in top-tier ML/AI venues (NeurIPS, ICML, CVPR, KDD, ACL, AAAI)
- Experience with LLMs (fine-tuning, distillation, RLHF, prompt engineering)
- Demonstrated ability to define and drive science roadmaps that influence product and business strategy
Our inclusive culture empowers Amazonians to deliver the best results for our customers. If you have a disability and need a workplace accommodation or adjustment during the application and hiring process, including support for the interview or onboarding process, please visit for more information. If the country/region you're applying in isn't listed, please contact your Recruiting Partner.
Job ID: 152203921
Skills:
Deduplication, Natural Language Processing, Java, Machine Learning, Deep Learning, Python, Information Retrieval, large-scale retrieval systems, multimodal ML, LLM-based policy reasoning, agentic architectures, vision-language models, video understanding, reinforcement learning, generative AI, embedding-based similarity, knowledge distillation
Skills:
Agentic systems, Decision systems, Multimodal understanding, Foundation models, Content Moderation
Skills:
Agentic systems, Decision systems, Large-scale classification, Multimodal understanding, Foundation models, Content Moderation
