About the job
In Amazon Advertising, we apply Machine Learning at massive scale to optimize programmatic advertising performance. The Demand Tech team owns response prediction and incrementality models that power bid optimization across Amazon DSP and Sponsored Display — determining how billions of ad impressions are valued and served daily across Amazon-owned properties, the open internet, and third-party exchanges.
Responsibilities
Own end-to-end response prediction — design and improve deep learning models for multi-task prediction (click, conversion, page view, incrementality) serving at inference latencies under 10ms at millions of TPS
Build and iterate on calibration mechanisms that keep prediction accuracy stable across rapidly shifting supply distributions
Integrate novel signals (OpenRTB features, customer behavioral sequences, supply quality feeds) into production models to improve optimization quality
Run online A/B experiments at scale, analyze results with statistical rigor, and translate offline gains into measurable business impact
Collaborate closely with engineers on model serving infrastructure (SageMaker, GPU inference, real-time feature stores) to deploy models efficiently at scale
Mentor scientists on the team and contribute to the broader Amazon ML science community through papers, conferences, and internal deep dives
Qualifications
Minimum
3+ years of building machine learning models for business application experience
PhD, or Master's degree and 6+ years of applied research experience
Experience programming in Java, C++, Python or related language
Experience with neural deep learning methods and machine learning
Preferred
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.