What does it take to build a foundation model that can forecast demand for hundreds of millions of products - including ones that have never been sold before?
At Amazon, our Demand Forecasting team is tackling one of the most ambitious challenges in applied time series research: building large-scale foundation models that generalize across an enormous and diverse catalog of products, geographies, and business contexts. This is not incremental modeling work. We are redefining what's possible in demand forecasting.
Our team operates at a scale that is unmatched in industry. We run experiments across millions of products simultaneously, pushing the boundaries of what foundation models can learn from vast, heterogeneous time series data. We are also exploring novel data generation techniques that augment our already unprecedented dataset - opening new frontiers in model generalization and forecasting for products with limited or no sales history.
The models you build here will ship to production and directly influence hundreds of millions of dollars in automated inventory decisions every week, labor plans for tens of thousands of employees, and Amazon's financial outlook. Beyond operational impact, this team contributes to the broader scientific community and advances the state of the art in time series foundation models.
If you are a scientist who wants to work at the frontier of time series research, at a scale no academic lab or startup can match, and see your work deployed to real-world impact - this is the team for you.
Key job responsibilities
- Design and run rigorous experiments at scale to evaluate and improve foundation model performance across hundreds of millions of products, geographies, and business verticals
- Lead the end-to-end lifecycle of forecasting models - from research and experimentation through production launch - including defining success metrics, obtaining stakeholder sign-off, and managing rollout
- Conduct online and offline labs to measure the real-world impact of forecast improvements beyond accuracy, including downstream supply chain, inventory, and financial outcomes
- Develop and deploy production-grade deep learning and statistical models using Python, Scala, SQL, and related tools
- Perform large-scale exploratory data analysis to uncover patterns, identify opportunities, and inform model development
- Translate complex research findings into clear insights and recommendations for technical and non-technical stakeholders at all levels
- Contribute to Amazon's scientific community and the broader research field through collaboration and publication in top-tier venues
A day in the life
No two days look the same, but most will involve some combination of deep technical work, cross-functional collaboration, and scientific thinking at a scale you won't find anywhere else.
You might start the morning reviewing the results of an experiment running across hundreds of millions of products - analyzing whether a new foundation model variant is improving generalization on cold-start items, or whether a novel data generation approach is meaningfully shifting forecast quality. You'll dig into the numbers, form a hypothesis, and design the next iteration.
Later in the day, you could be in a stakeholder review, walking business and engineering partners through a set of launch metrics - explaining not just forecast accuracy, but the downstream supply chain and financial impact your model is driving. Getting a model to production at Amazon requires rigor: you'll define success criteria, run online and offline labs to validate real-world impact, and build the case for sign-off across technical and business stakeholders.
You'll write code - Python, Scala, SQL - to process and analyze data at a scale most scientists never encounter. You'll collaborate closely with scientists, engineers, and business teams, and contribute to research that has a real chance of being published and advancing the field.
The work is hard, the problems are unsolved, and the impact is immediate. If you want to do research that ships - this is where you do it.
BASIC QUALIFICATIONS
- 1+ years of data querying languages (e.g. SQL), scripting languages (e.g. Python) or statistical/mathematical software (e.g. R, SAS, Matlab, etc.) experience
- 2+ years of data/research scientist, statistician or quantitative analyst in an internet-based company with complex and big data sources experience
- Bachelor's degree
PREFERRED QUALIFICATIONS
- Knowledge of machine learning concepts and their application to reasoning and problem-solving
- Experience applying quantitative analysis to solve business problems and making data-driven business decisions
The base salary range for this position is listed below. Your Amazon package will include sign-on payments and restricted stock units (RSUs). Final compensation will be determined based on factors including experience, qualifications, and location. Amazon also offers comprehensive benefits including 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, and parental leave. Learn more about our benefits at https://amazon.jobs/en/benefits.
USA, WA, Bellevue - 108,300.00 - 160,000.00 USD annually
