Amazon
Data Scientist , Reactive Transfers
$136K–$184K
Austin, TX · Full-time · Posted Oct 6, 2026
- Experience
- 2+ years as a data scientist; 3+ years in querying/scripting languages or statistical/mathematical software and 3+ years in machine-learning/statistical modeling and data-analysis methods; 1+ years each guiding/coaching researchers, working with or evaluating AI systems, and contributing to textbooks, research papers, or educational content.
- Education
- Master's degree in Science, Technology, Engineering, or Mathematics (STEM), or experience working in STEM.
About the job
Develops and evaluates data science and machine-learning solutions to improve Amazon’s supply-chain and fulfillment decisions, using large datasets and communicating findings to business stakeholders.
What you'll do
- Analyze supply-chain data and develop or improve statistical and machine-learning models through experiments and tuning.
- Define expected model behavior, identify and investigate outliers, and recommend fixes.
- Communicate research, insights, and recommendations to business customers and cross-functional partners.
What they ask for
- 2+ years of data scientist experience.
- 3+ years of experience with data-querying or scripting languages, or statistical/mathematical software; 3+ years with machine-learning/statistical modeling, data-analysis tools, techniques, and performance parameters.
- 1+ years of experience guiding and coaching researchers; 1+ years working with or evaluating AI systems; 1+ years creating or contributing to mathematical textbooks, research papers, or educational content.
- Master's degree in STEM, or experience working in STEM; experience applying theoretical models in an applied environment.
Nice to have
- Ph.D. in STEM.
- Experience defining and creating benchmarks for assessing GenAI model performance.
Pay
136,000.00 - 184,000.00 USD annually (base salary)
Tools
SQL, Python, R, SAS, Matlab, Perl
AI focus
Experience working with or evaluating AI systems is required; GenAI model-performance benchmarking is preferred.
Model-written summary · the original posting is authoritative