Do you want to teach machines what makes an ad creative truly effective? The Creative X team within Amazon Advertising is building AI-driven systems that understand and evaluate ad creatives - images, videos, audio, and text - so advertisers of every size can deliver high-quality experiences to customers. This role sits in our creative understanding charter, where you will design models that identify what a creative contains, assess whether it meets quality and brand requirements, and measure effectiveness across multiple dimensions with measurable precision and recall.
We invest in multimodal large language models (LLMs), vision-language models, computer vision, representation learning, and evaluation science including automated-judge calibration and preference learning from human feedback. A key challenge you will tackle is making these capabilities work at massive scale through techniques like model distillation, enabling real-time understanding and evaluation across billions of ad impressions. You will join a collaborative team of applied scientists, product managers and software developers, ship advertiser-facing features, publish research, and help set the scientific direction for creative quality at Amazon Advertising.
Key job responsibilities
- Design and build multimodal models that extract semantic metadata from ad creatives - identifying objects, scenes, text overlays, brand elements, and audio cues - to power downstream understanding and evaluation systems.
- Develop evaluation frameworks that measure creative quality and effectiveness across defined dimensions and attributes, detecting production defects and assessing brief compliance with measurable precision and recall.
- Apply distillation and compression techniques to scale large vision-language and LLM-based models into efficient, production-ready systems that operate across billions of daily ad impressions.
- Calibrate automated evaluation models against human judgment through preference learning and judge-alignment methods, continuously improving agreement with expert assessors.
- Collaborate with engineering and product partners to experiment quickly and move models from experimentation to production, and mentor scientists on the team through design reviews, code reviews, and knowledge sharing.
BASIC QUALIFICATIONS
- PhD, or Master's degree and 6+ years of applied research experience
- Experience programming in Java, C++, Python or related language
- Experience building and maintaining fully automated large-scale ML systems that self-improve
PREFERRED QUALIFICATIONS
- Experience with modeling tools such as R, scikit-learn, Spark MLLib, MxNet, Tensorflow, numpy, scipy etc.
- Experience with large scale distributed systems such as Hadoop, Spark etc.
The base salary range for this position is listed below. Your Amazon package will include sign-on payments and restricted stock units (RSUs). Final compensation will be determined based on factors including experience, qualifications, and location. Amazon also offers comprehensive benefits including health insurance (medical, dental, vision, prescription, Basic Life & AD&D insurance and option for Supplemental life plans, EAP, Mental Health Support, Medical Advice Line, Flexible Spending Accounts, Adoption and Surrogacy Reimbursement coverage), 401(k) matching, paid time off, and parental leave. Learn more about our benefits at https://amazon.jobs/en/benefits.
USA, WA, SEATTLE - 167,100.00 - 226,100.00 USD annually