Data Scientist Iii, Customer Strategy

Amazon Amazon · Big Tech · NY +1 · Data Science

Data Scientist III role focused on building and deploying machine learning models for customer segmentation, personalization, and optimization in a consumer retail setting. The role involves applying causal inference, LLMs, and advanced optimization techniques to drive evidence-based decisions and improve customer experiences across various channels. Collaboration with engineering for data pipelines and serving infrastructure is key, with a focus on translating analytical results into actionable business strategies.

What you'd actually do

  1. Design, build, and iterate on customer segmentation models that drive product recommendations, content ranking, intent detection, and customer-specific experiences on site, in email, and in push notifications across Shopbop and Zappos.
  2. Apply advanced optimization techniques — including uplift modeling, to improve real-time decisioning across marketing, digital, and channel experiences.
  3. Apply causal inference methods grounded in econometric and machine learning frameworks, including EconML, DoWhy, and CausalML, to estimate the true incremental lift of personalization strategies and marketing interventions through techniques such as double machine learning, meta-learners (T-learner, S-learner, X-learner), and targeted maximum likelihood estimation.
  4. Build and maintain predictive models for customer preferences and individualized treatment effect models that inform business strategy and investment decisions.
  5. Collaborate with Engineering to build scalable data pipelines, feature stores, and real-time serving infrastructure that support ongoing model development and experimentation.

Skills

Required

  • SQL
  • Python
  • R
  • SAS
  • Matlab
  • statistical models
  • multinomial logistic regression
  • machine learning
  • causal inference
  • EconML
  • DoWhy
  • CausalML
  • double machine learning
  • meta-learners
  • targeted maximum likelihood estimation
  • uplift modeling
  • data pipelines
  • feature stores
  • serving infrastructure

Nice to have

  • AWS QuickSight
  • Tableau
  • R Shiny
  • generative AI tools
  • prompting
  • evaluation practices

What the JD emphasized

  • customer segmentation models
  • personalization
  • causal inference
  • machine learning
  • LLMs
  • optimization techniques
  • predictive models
  • individualized treatment effect models
  • data pipelines
  • feature stores
  • serving infrastructure
  • deploy data science models
  • production

Other signals

  • customer segmentation models
  • personalization
  • causal inference
  • machine learning
  • LLMs
  • optimization techniques
  • predictive models
  • individualized treatment effect models
  • data pipelines
  • feature stores
  • serving infrastructure
  • deploy data science models
  • production