Amazon Applied Scientist II EC2 Onhand Packing Science 

United States, Washington, Seattle 

698645275

Key job responsibilities
We are seeking an expert with a strong background in mathematical optimization with excellent modeling skills, with experience using both exact and heuristic methods. The domain requires a tight collaboration with engineering teams, to ensure we are able to quickly move from modelling and prototyping to full scale production usage. You should be confirmable with the needed statistical approaches for production A/B tests as well as have a solid understanding experimental design more broadly. Being successful in the role requires having the scientific breadth to understand the interactions between different phases of a project from data analysis through to production, including resolving issues after rollout.You will be hands-on with the mathematical modeling and implementation, and will also contribute to the design of the engineering system with the scalability, extensibility, maintainability, and correctness of the optimization engine in mind. You will review approaches by other scientists and engineers in terms of business relevance, technical validity, engineering / science interface, and computational performance. You will mentor and lead junior scientists by example. Communicating your results to guide the direction of the business and working with software development teams to implement your ideas in code is key to success. You will write technical, and less frequently, business documents that influence engineering investments and business direction. Collaborating with other scientists, software engineers, and product managers, you will develop creative, novel, and data-driven approaches to improve our existing cloud compute offerings and define new ones in a fast-paced and quickly changing environment, improving the experience of our customers and impacting the bottom line of EC2.

- PhD in operations research, applied mathematics, computer science, or similar quantitative field.
- 2+ years of relevant academic research or industry experience in developing algorithms.
- Publications in refereed academic journals or conference.
- In-depth knowledge of exact, approximation algorithms, and heuristic methods for solving difficult optimization problems.
- Ability to productionize models end-to-end.
- Experience in prototyping and developing software in traditional programming languages (Python, C++, Java etc.).
- Comfortable with SQL and experience with very large-scale dataset.
- Excellent writing skills to document the models and analyses and for presenting business cases with results/conclusions in order to influence important decisions.