Senior Applied Scientist, GEOS Science Team

This role sits in Business Innovation and AI (BIA), part of Global Ops Enterprise Services (GOES), and works on the Real Estate Optimization System (REOS). REOS decides where Amazon builds. It models the supply chain from fulfillment centers through sort centers and delivery stations to the customer as a time-space network over a 20-year horizon, and ranks candidate sites on total network cost — transportation, labor, real estate and unmet demand — rather than on rent alone. Alongside the core engine we are building a family of AI tools that source candidate properties from across the market, assemble and validate the data behind a study, and generate the cost briefs that explain a recommendation, and we are extending the same optimization to zip-code-level Sub Same Day siting. You will work alongside scientists, and with partner teams in GOES Portfolio Management, Capacity Planning and Finance. We are looking for an Applied Scientist to own the optimization models behind REOS. You will formulate the models that decide where Amazon builds, engineer the solution methods that make them tractable at network scale, lead the initiatives that extend them, and take the recommendations to the Development Managers and Finance partners who act on them. The work is hands-on and close to the decision: your models become the basis for real capital commitments. The people who make those commitments are your customers. You will work with Development Managers, Capacity Planning teams and Finance partners to understand what they need, build and run the model that answers it, deliver the result, and explain what it recommends in terms they can act on and defend. Reporting to the Science Manager, you will own the technical approach on the models you build and mentor the scientists working alongside you. Key job responsibilities • Formulate and solve large-scale optimization models — linear, mixed-integer, network flow and stochastic or robust variants — that recommend site locations and network topology at lowest total cost. • Select and engineer the solution methodology — decomposition, column generation, cutting planes, relaxations and heuristics — so models solve reliably at production scale and within business timelines. • Own the modeling assumptions and the cost structure behind them, and make the trade-offs explicit and defensible to Finance. • Implement and productionize models in Python against commercial and open-source solvers, with the data pipelines and validation that keep them trustworthy, and maintain them as the network and the data change. • Engage customers directly: elicit the real problem, translate it into a formulation, run scenarios with them, deliver results, and explain what the model recommends and why. • Lead initiatives end to end, apply AI tooling to the science workflow, and raise the bar for the scientists around you through design review and mentoring. A day in the life You might start your morning reviewing experiment results and refining a machine learning model before syncing with engineers on a production integration plan. After lunch, you could be whiteboarding a new algorithmic approach with colleagues from product and science teams, then drafting a research document that outlines your methodology and findings. You will regularly present your work to leadership, participate in science reviews, and stay connected to the latest research in your field to inform your next initiative. Amazon Benefits: Amazon offers a full range of benefits that support you and eligible family members, including domestic partners and their children. Benefits can vary by location, the number of regularly scheduled hours you work, length of employment, and job status such as seasonal or temporary employment. The benefits that generally apply to regular, full-time employees include: 1. Medical, Dental, and Vision Coverage 2. Maternity and Parental Leave Options 3. Paid Time Off (PTO) 4. 401(k) Plan If you