Data Scientist II: Forecasting & Optimization for NASC

Amazon; Bellevue (WA); On-site; USD 136,000 - 184,000; Full time

Benefits offered by this job

Health insurance

RSUs

401(k) matching

Paid time off

Parental leave

Job summary

Amazon NASC & TOM Science team seeks a Data Scientist II to own end-to-end machine learning and optimization projects that inform North America Sort Center and Transportation Operations planning. You will frame problems with business partners, build models, validate, deploy to hosting platforms, and monitor performance in production environments.

You will work with Engineering, Product, and stakeholders to translate operational pain points into measurable model outcomes, covering forecasting,

Qualifications

• 2+ years of data scientist experience.
• 3+ years of data querying languages (SQL) or scripting (Python) experience.
• 3+ years of ML/statistical modeling and data analysis with applied results.

Responsibilities

• Design and implement ML and optimization solutions (forecasting, MIP/LP, DL).
• Drive end-to-end delivery from data exploration to deployment and monitoring.
• Develop modeling patterns for forecasting and optimization.
• Build validation, backtesting, and monitoring pipelines; reduce leakage and bias.
• Define model performance metrics tied to business outcomes.
• Collaborate with Data Engineering and Software teams to productionize models.
• Communicate findings clearly to stakeholders and senior leadership.

Skills

SQL

Python

Statistical modeling

Education

Master's in STEM

Tools

AWS services (S3, Redshift, SageMaker, EMR, Kinesis, Lambda, EC2)

Job description

Amazon NASC & TOM Science team seeks a Data Scientist II to own end-to-end machine learning and optimization projects that inform North America Sort Center and Transportation Operations planning. You will frame problems with business partners, build models, validate, deploy to hosting platforms, and monitor performance in production environments.

You will work with Engineering, Product, and stakeholders to translate operational pain points into measurable model outcomes, covering forecasting,