SCOT Network Topology science team focuses on research areas of forecasting and optimization models that determine Amazon outbound network design as we transition to relying on our internal carrier network and accelerate one-day delivery speed. We utilized models from various science disciplines such as: Mixed Integer , Random Forest (or other ML techniques), probabilistic model, stochastic economic analysis, to name a few.
In addition to network, we also use and techniques to evaluate new facilities recommendation for long term estimates, We use to approximate the network, and simulation of how our choices will perform.
We are looking for a Research Scientist who has a knowledge of analyzing fulfillment data using statistics and machine learning models . Those who are strong in space should have a breadth of other ML experience in a production environment using techniques. This role will focus on expanding our reach to analyze various fulfillment and for Amazon's network worldwide.
To help describe some of our challenges, we created a short video about at Amazon - http://bit.ly/amazon-scot
A day in the life
A scientist in the team is expected to do 4 mixes of works: (i) Research development - this can be new model or existing model feature enhancement (40-50%), (ii) Ad-hoc study - pulling data or create simple prototype to answer a business questions (20-30%) and (iii) Operational (20-30%) - this includes meeting with stakeholders for feedback, learning new things, or team meetings
About the team
SCOT Network Topology science team focuses on research areas of forecasting and optimization models that determine Amazon outbound network design as we transition to relying on our internal carrier network and accelerate one-day delivery speed. There are various strategic questions the team is attempting to answer, such as: what is the impact of placement on outbound cost and delivery speed? What is the optimal network design given capacity constraints? How can we forecast accurately fulfillment pattern for different customer clusters?. If you are interested in diving into a multi-discipline, high impact space this team is for you.
Our team is a mixture of Software Engineers, Operations Research Scientists, Applied Scientists, Business Intelligence Engineers and Product Managers.
PhD in a quantitative field such as Mathematics, Statistics, Physics, Engineering, Computer Science, , Economics and 3+ years of industry experience OR MS or greater in a quantitative field and 6+ years of experience (after graduation)
Strong coding and problem-solving skills in at least one programming such as Python, Java, SQL, etc.
Familiar with AWS environment, such as S3, Sage Maker and other
Experience with linear or model to make big-impact decision making. Familiar with using CPLEX/XPRESS/Gurobi
Sound theoretical understanding of broad concepts, with and demonstrable expertise in at least one topic or analytics model in
Experience with fully automated training (e.g. automatic re-training, automatic testing) on techniques such as Random Forest, Regression (Linear), Time-Series (ARIMA) and Neural network (LSTM, CNN) using 1+GB datasets.
Experience with solving business problems with high-impact decisions (> $50M), consulting senior management on benefits of new design based economic ROI analysis using techniques and explanation of models and its impact
Experience with fine tuning and building large-scale model using XPRESS/Mosel, Gurobi or CPLEX, decomposition technique or designing heuristics.
Experience writing production-quality code using collaborative process such as Git and AWS with other scientists or software engineers.
Amazon is committed to a diverse and inclusive workplace. Amazon is an equal opportunity employer and does not discriminate on the basis of race, national origin, gender, gender identity, sexual orientation, protected veteran status, disability, age, or other legally protected status. For individuals with disabilities who would like to request an accommodation, please visit https://www.amazon.jobs/en/disability/us.