Applied Scientist, Prime Video - Marketing Science

Amazon Amazon · Big Tech · Seattle, WA · Applied Science

Applied Scientist role focused on building sequential decision-making frameworks and predictive models for marketing resource allocation and budget optimization within Prime Video. The role involves automating and scaling modeling infrastructure and collaborating with business and finance teams to translate business questions into statistical and optimization problems. Requires experience in machine learning and algorithm development for business applications.

What you'd actually do

  1. Build sequential decision-making frameworks (e.g., MDPs, multi-armed bandits, dynamic programming) to optimize marketing resource allocation over time under uncertainty.
  2. Create predictive models to forecast marketing efficiency and aid strategic budget allocation across channels, campaigns, and markets.
  3. Design and analyze geo-level and regional hold-out experiments to validate model predictions and establish ground truth.
  4. Automate and scale our modeling infrastructure to improve efficiency and expand coverage across use cases and geographies.
  5. Collaborate with leaders across business and finance teams to translate business questions into well-posed statistical and optimization problems.

Skills

Required

  • Master's degree or above in Engineering, Computer Science, Machine Learning, Operations Research, Statistics, or related fields
  • Experience programming in Java, C++, Python or related language
  • Experience building machine learning models or developing algorithms for business application

Nice to have

  • Experience implementing algorithms using both toolkits and self-developed code
  • Have publications at top-tier peer-reviewed conferences or journals
  • PhD in Engineering, Computer Science, Machine Learning, Operations Research, Statistics, or related fields

What the JD emphasized

  • building machine learning models or developing algorithms for business application

Other signals

  • Build sequential decision-making frameworks (e.g., MDPs, multi-armed bandits, dynamic programming) to optimize marketing resource allocation over time under uncertainty.
  • Create predictive models to forecast marketing efficiency and aid strategic budget allocation across channels, campaigns, and markets.
  • Automate and scale our modeling infrastructure to improve efficiency and expand coverage across use cases and geographies.