Applied Scientist, Amazon Live Data Engineering, Sciences and Analytics (DESA)
Core
Building production-grade causal measurement, content intelligence, and ranking signals for Amazon Live's shoppable video platform to prove video value to brands and influence product decisions.
Role type
Applied Scientist (Causal Inference & Content Intelligence)
Builds
Causal attribution models, brand lifecycle models, multimodal/generative content intelligence, ranking features, and A/B experimentation frameworks.
Domain
E-commerce / Live Video / Digital Advertising
Deliverable
production ML models
Required skills
Causal inference methods, A/B testing design, large-scale distributed systems, machine learning fundamentals, multimodal/generative modeling, Python/Java/C++ programming
Preferred skills
Search advertising experience, video/image processing/computer vision, customer lifecycle marketing, Amazon internal tools (Bedrock, SageMaker)
Technologies
PyTorch, JAX, Hadoop, Spark, Bedrock, SageMaker, Redshift
Responsibilities
Design and deploy causal attribution models (incrementality testing, multi-touch), build brand lifecycle and campaign optimization models, design and run A/B experiments, develop multimodal models for content intelligence, build ranking and personalization features, own full lifecycle from research to production deployment.
Seniority
Mid-Senior, hands-on IC
