Sr. Applied Scientist, PXT Central Science

Amazon • Arlington, VA

11d

About The Position

Every day, hundreds of thousands of Amazon associates show up to fulfill the
promise we make to our customers. Behind the workforce decisions that support
them — staffing, retention, scheduling, development — there should be science
that doesn't just describe what happened, but explains why it happened and
predicts what comes next. That's the work we do.

PXT Central Science (PXTCS) is Amazon's internal research organization dedicated
to bringing scientific rigor to people and workforce decisions at global scale.
Our team sits within the part of PXTCS that focuses on Amazon's Tier 1 hourly
populations — the associates at the heart of Amazon's operations. We are a
multidisciplinary group of 15 economists, data scientists, data engineers, and
research scientists united by a single mission: to transform complex operational
challenges into actionable insights through rigorous causal analysis and
predictive modeling that empowers data-driven workforce decisions.

We are building something new — causal predictive models that go beyond
traditional forecasting. Our models don't just tell leaders what will happen;
they reveal why it will happen and what levers they can pull to change the
outcome. This is the frontier where causal inference meets modern machine
learning, and we need a scientist who can help us push it forward.
As a Senior Applied Scientist on this team, you will be the connective tissue
between innovative research and real-world impact. You will work
shoulder-to-shoulder with economists who deeply understand the causal mechanisms
driving workforce dynamics and data scientists who know the operational
landscape — and you will bring the technical creativity to expand what's
possible. That means writing production-quality code that our partner
engineering teams can implement into decision-making tools. It means exploring
novel feature spaces — large language models, computer vision, and other
emerging techniques — to unlock signal that traditional approaches miss. And it
means doing all of this with the scientific rigor that causal claims demand.

This role is built for someone who is entrepreneurial and energized by ambiguity
— someone who sees a prototype model and immediately starts thinking about how
to make it robust, scalable, and impactful. You will not just advance your own
work; you will elevate the scientists around you. We are looking for a strong
technical individual contributor who is passionate about developing peers
raising the bar across disciplines, and who sees a future path into a science
manager position.

If you want to do science that directly shapes how Amazon supports its workforce
— not in theory, but in production systems that leaders use to make better
decisions every day — we'd love to talk.

Requirements

• PhD, or Master's degree and 6+ years of applied research experience
• 5+ years of building machine learning models for business application experience
• Experience programming in Java, C++, Python or related language
• Experience with modeling tools such as R, scikit-learn, Spark MLLib, MxNet, Tensorflow, numpy, scipy etc.

Nice To Haves

• PhD in econometrics, statistics, industrial engineering, operations research, optimization, data mining, analytics, or equivalent quantitative field
• Experience with neural deep learning methods and machine learning
• Experience in causal modeling like graphical models, causal Bayesian network, potential outcomes, A/B testing, experiments, quasi-experiments, and data science workflows
• Experience in taking a product from conception & definition phase through engineering design and taking it to market
• Experience working with emerging technologies
• Experience in mentoring, leading and coaching

Responsibilities

• Design and build causal predictive models that move beyond correlation — developing systems that forecast workforce outcomes and identify the actionable drivers behind them, enabling leaders to intervene before problems materialize
• Pioneer novel feature engineering by bringing creative approaches from LLMs, computer vision, and other emerging techniques into the causal modeling pipeline, unlocking signal that traditional econometric and tabular methods miss
• Write production-quality science code that your partner engineering team can implement directly into operational decision-making tools — your work must be clean, well-documented, and built to scale
• Bridge disciplines by translating between economists, data scientists, and engineers — synthesizing causal rigor with ML innovation to produce models that are both scientifically defensible and operationally useful
• Design and execute experiments to validate causal claims and model performance, establishing evaluation standards that the team and stakeholders trust
• Develop and elevate peers across the team — mentoring scientists in adjacent disciplines, sharing technical knowledge, and raising the collective bar on modeling and engineering practices
• Present findings to senior leadership, distilling complex causal and predictive insights into clear recommendations that drive workforce strategy for Amazon's Tier 1 hourly populations.

Benefits

• health insurance (medical, dental, vision, prescription, Basic Life & AD&D insurance and option for Supplemental life plans, EAP, Mental Health Support, Medical Advice Line, Flexible Spending Accounts, Adoption and Surrogacy Reimbursement coverage)
• 401(k) matching
• paid time off
• parental leave

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level