remote
Applied Scientist, Traffic Quality - Amazon.com
Software Engineer
Lead the development of scalable ML models to detect and mitigate invalid traffic across Amazon Ads platforms, leveraging Python, AWS, and big‑data technologies to protect advertiser trust at petabyte scale.
About the role
Key Responsibilities
- Design, prototype, and deploy production‑grade machine learning models that identify sophisticated invalid traffic patterns across desktop, mobile, and connected TV.
- Engineer end‑to‑end data pipelines on AWS (S3, Glue, Redshift, SageMaker) to ingest, process, and analyze petabyte‑scale clickstream data.
- Collaborate with cross‑functional teams to define feature sets, evaluate model performance, and iterate on solutions that balance detection accuracy with latency constraints.
- Conduct rigorous experimentation, A/B testing, and statistical analysis to validate model impact on advertiser trust and revenue.
- Document methodologies, maintain reproducible research notebooks, and present findings to stakeholders.
Requirements
- 5+ years of experience building large‑scale ML systems in a production environment.
- Proficiency in Python, SQL, and AWS services (SageMaker, Glue, Redshift, EMR).
- Strong background in supervised/unsupervised learning, feature engineering, and model deployment.
- Experience with big‑data processing frameworks (Spark, Flink) and time‑series analysis.
- Excellent communication skills and a track record of translating complex data insights into actionable business decisions.
Skills
pythonmachine learningaws