Applied Scientist II

Saved job description

Every year, Amazon's operations depend on hiring hundreds of thousands of hourly associates across the globe, and your work will determine how efficiently and effectively that happens. We are looking for a curious, resourceful Applied Scientist to tackle multi-objective optimization problems worldwide. In this role, you will build mathematical models and AI/ML systems that optimize recruitment processes under real-world cost, capacity, and regulatory constraints across North America and beyond. This is a particularly exciting time to join because we are expanding into new countries, building AI-native capabilities, and beginning to externalize our solutions so that teams beyond our own can benefit from the tools we create. You will partner with economists, machine learning engineers, and analytics engineers to turn theory-based decision models into production systems that improve candidate experience and business outcomes at a scale few teams can offer.

Key job responsibilities

- Design and implement optimization algorithms and mathematical models that determine optimal job posting timing, candidate management strategies, and recruitment coordination across multiple countries and markets.

- Develop and maintain conversion models, demand forecasting systems, and multi-objective optimization frameworks using machine learning, survival analysis, and causal inference techniques.

- Plan and run controlled experiments to validate new recruitment mechanisms, measure their impact, and iterate on model performance using real-world data.

- Write production-quality code in Python, applying optimal data structures and algorithms while making practical tradeoffs between complexity, memory efficiency, and maintainability.

- Collaborate with generation, hiring event, and coordination teams to integrate your models into automated hiring pipelines, and contribute to the team's scientific agenda through peer reviews, mentorship, and publications.

A day in the life

You might start your morning reviewing model performance dashboards and conversion metrics, then refine a stochastic dynamic programming model that balances hiring demand against cost constraints for an upcoming peak period. After lunch, you could be designing an A/B test for a new candidate management strategy or pairing with another scientist to incorporate local market regulations into an optimization framework. You will regularly present findings in science reviews and help shape the team's research roadmap.

About the team

Our team sits within Amazon's operations organization and owns the science behind converting job applicants into hired associates at a global scale. We support leadership teams across North America and beyond, building the optimization and ML systems that keep fulfillment centers staffed and ready to serve customers. Right now we are investing in AI-native automation, expanding our models to additional countries, and making our tools available to partner teams across the organization. The team is a multidisciplinary group of scientists, economists, machine learning engineers, and analytics engineers who collaborate closely to solve problems that span economic theory, statistical modeling, and large-scale engineering. We value rigorous experimentation, collaborative problem-solving, and a shared commitment to improving both business outcomes and the candidate experience.

Requirements: - 3+ years of building models for business application experience

- Experience programming in Java, C++, Python or related language

- Experience building machine learning models or developing algorithms for business application

- Experience in designing experiments and statistical analysis of results

- Experience in solving business problems through machine learning, data mining and statistical algorithms

- Experience in state-of-the-art deep learning models architecture design and deep learning training and optimization and model pruning