Software Development Engineer, Efficiency & Performance, Alloy

In eCommerce, we operate in an ultra-low-margin business. Our ability to operate efficiently makes the difference between profit and loss, or between prices a customer is happy or displeased with. It is at the core of our business DNA, enshrined in both the Leadership Principle of Frugality and the Amazon flywheel (lower cost structure). The costs of running the Amazon website are now measured in billions of dollars, and it is our collective responsibility to ensure that we are realizing an economy of scale, rather than having scale kill our economy.

The Alloy team is a central efficiency team which drives cost savings for all service teams within Amazon via efficient use of AWS resources as they build and operate their services. This team is special in 3 ways (1) business impact - we have proven records to save cost by hundred million dollars annually. We have earned trust and reputation from service teams, partner teams (business and technical), and senior leadership (2) technical complexity - our system is not a single product but the whole Amazon. We create central efficiency solutions which save costs for thousands of internal services with minimal or zero efforts from their engineers; (3) professional network - we work with a group of Principal Engineers and Distinguished Engineers closely. Working with brilliant people helps you grow your career.

We are standing up a new central Performance Engineering team to expand our scope from hardware utilization (unused resources) to software optimization (used resource), via profiling with central fixes, proper upgrade of library, central scheduling from resource pools, etc., In addition, we are expanding our scope from compute to storage resources, such as auto migration to cheaper storage classes, removal of duplicates and unused data, applying proper data retention policy, logging with proper compression format, etc. We are also building a large resource sharing infrastructure to maximize the utilization of scarce GPU resources. While a lot of these are common techniques for individual teams, how to create a central, safe, and scalable solution to reduce their undifferentiated efforts is the amazing part of this team! In this team, we never lack of innovations! As we are realizing savings from the top opportunities this year, we are already exploring new saving opportunities for the next year! We are always good at discovering green fields! Sky is the limit!

As a Software Development Engineer on this team, you will work with software teams across Amazon to learn, innovate, and lead a team of engineers to enhance the performance and efficiency of software solutions in all areas of Amazon's software stack. You will search for, uncover, and improve software and architecture inefficiencies in the vast Amazon ecosystem that will result in millions of dollars of savings in infrastructure costs. You will invent, define, collect, aggregate, and analyze various efficiency metrics, and use them to set and drive efficiency goals for Amazon and service teams. You will automate the measurement with a data-driven approach without minimal manual efforts. You will generate reports for the potential and actual cost savings to our finance partners and senior leadership regularly. This is a great opportunity to be involved from the beginning in this new efficiency initiative and be part of the program that will make a lasting impact at Amazon!

In this role, you will:

* Solve performance and efficiency problems that manifest at scale.
* Design metrics and measure performance and cost efficiency of services in Amazon's ecosystem.
* Collaborate with service teams to identify inefficiencies, and design and implement solutions.
* Design and develop highly available components and profiling tools.
* Lead and mentor a team of engineers.