At Amazon, our SCOT Labs team owns and operates the experimentation platform that powers randomized controlled trials (RCTs) across Supply Chain Optimization Technologies (SCOT). We are the scientific gatekeepers for policy updates that govern how Amazon buys, stores, and moves billions of units of inventory worldwide. This is not traditional A/B testing: we are building the infrastructure and methodology to causally evaluate complex and interconnected supply chain interventions. Our platform runs experiments that span millions of products and hundreds of fulfillment nodes simultaneously, measuring the real-world impact of policy changes on inventory health, customer experience, and operational cost. We are also advancing the science of causal inference in supply chain settings by developing novel approaches to treatment effect estimation, interference modeling, and emulation techniques that allow us to assess policy impact faster and more accurately than ever before. The experiments you design and the methods you build here will directly determine which policies ship to production. These decisions influence hundreds of millions of dollars in weekly inventory investments, labor allocation for tens of thousands of associates, and Amazon's overall supply chain efficiency. Beyond operational impact, this team pushes the frontier of causal experimentation methodology and contributes to the broader scientific community with publications at top venues. If you are a scientist who wants to shape how one of the world's largest supply chains makes decisions — solving causal inference challenges in real-world settings no academic lab or startup can replicate — this is the team for you. Key job responsibilities - Partner with customer teams to design rigorous large-scale experiments (such as randomized controlled trials and quasi-experiments) to evaluate policy updates and model improvements across millions of products, hundreds of fulfillment nodes, and diverse business contexts - Build and maintain production-grade experimentation infrastructure and analytical tools using Python, SQL, Scala, and related technologies - Advance causal inference methodology for supply chain settings, including treatment effect estimation, interference modeling, and emulation techniques that accelerate policy evaluation. Ensure these methods are supported by both theoretical foundations and empirical evidence. - Translate complex research findings into clear insights and actionable recommendations for technical and non-technical stakeholders at all levels. - Contribute to Amazon's scientific community and the broader external research field through collaboration and publication in top-tier venues. A day in the life You might start the morning digging into results from a randomized trial running across millions of products, deciding whether the estimator you built last month is holding up and where it breaks. By the afternoon you could be designing the next experiment with a partner team where clean randomization is not available, because treated and control units draw on the same fulfillment network and inventory pools, and you need an approach that stays credible despite the spillover. The role sits between developing methods and shipping them. You will build estimators and experiment designs for problems without textbook answers, prototype them in Python on real supply chain experiment data, and then put the ones that work into the platform other SCOT scientists use, writing Python, Scala, and SQL at a scale most scientists never encounter. You will also build supply chain emulation systems that replicate inventory dynamics end to end, so a design can be stress tested before it touches the real network. The problems are open, the data is unmatched in scale, and the methods you build will shape how scientists across SCOT think about causal inference and move billions of dollars in inventory decisions. If you want to do measurement and experimentation research that actually ships, this is where you do it. About the team The Forecasting and Labs Science team sits at the heart of Amazon's supply chain, building the science that determines what products are available, when, and at what cost for hundreds of millions of customers around the world. Our mission spans two deeply connected frontiers: pushing the boundaries of large-scale time series forecasting through foundation models that generalize across an enormous and diverse catalog of products, and building the experimentation and causal inference methodology that rigorously evaluates whether supply chain policy changes should ship to production. We are a team of scientists who care deeply about both research rigor and real-world outcomes. We don't just publish: we ship. And we don't just ship: we measure, iterate, and raise the bar. On the forecasting side, we build foundation models at a scale unmatched in industry, running experiments across millions of products and exploring novel data generation techniques that open new frontiers in model generalization. On the experimentation side, we design and run randomized controlled trials across hundreds of fulfillment nodes, advance causal inference in settings where interference is unavoidable, and build supply chain emulation systems that can evaluate policy changes in hours rather than months. Our work spans the full lifecycle: from foundational research and large-scale experimentation to production deployment and downstream impact measurement across supply chain, inventory, and financial planning.
