Senior Machine Learning Engineer, Sustainability Science and Innovation
Amazon · Seattle, Washington, USA ·
- Seniority
- Senior
- Category
- ML ai
Join us at the forefront of Amazon's sustainability initiatives to work on environmental and social advancements to support Amazon's long term worldwide sustainability strategy. At Amazon, we're working to be the most customer-centric company on earth. To get there, we need exceptionally talented, bright, and driven people.
The Worldwide Sustainability (WWS) organization capitalizes on Amazon’s scale and speed to build a more resilient and sustainable company. We manage our social and environmental impacts globally, and drive solutions that enable our customers, businesses, and the world to become more sustainable.
We are looking for an engineer to build the software and infrastructure that help scientists develop, evaluate, and deploy machine learning models for sustainability applications. You will work directly with applied scientists, turning production-grade science artifacts into reliable systems and building reusable tools that reduce the engineering work required for each new model.
The work spans research workflows and production systems: preparing data for experiments, automating training and evaluation, and making models available through dependable interfaces. A central design challenge is deciding what the platform should handle consistently and where researchers need flexibility. Good abstractions let scientists change a model or transformation without rebuilding the surrounding infrastructure, while keeping the underlying system accessible when something needs to be understood or fixed.
You will also develop interfaces and workflows that support AI-assisted engineering. This includes making platform capabilities understandable to coding agents and building checks that verify their changes before those changes reach production.
Key job responsibilities
- Build reusable data and ML infrastructure. Develop libraries, configuration-driven interfaces, and infrastructure-as-code components for data processing, model training, and inference.
- Build tools for repeatable experiments and model comparisons, preserving the data, code, and configuration needed to investigate results.
- Develop and evaluate AI-assisted workflows. Create tooling that helps coding agents use the platform correctly. Assess their output through automated checks and appropriate review, and measure whether the workflows improve engineering productivity without weakening reliability.
- Support adoption and evolution. Work with platform users to identify recurring needs, document design trade-offs, and maintain compatibility as shared components change.
About the team
Our team brings together applied scientists and engineers to develop models and software for sustainability applications. We work across research and production, building both scientific methods and the systems needed to use them reliably. That combination keeps platform development close to the practical needs of researchers and the people who depend on their models.
We build reusable tools so that each new research project does not require its own infrastructure. The engineering challenge is to make experimentation easier while preserving the checks and evidence needed to understand results and operate models in production. Scientists and engineers work together on those decisions, including where a shared approach helps and where a scientific problem requires something different.
The Worldwide Sustainability (WWS) organization capitalizes on Amazon’s scale and speed to build a more resilient and sustainable company. We manage our social and environmental impacts globally, and drive solutions that enable our customers, businesses, and the world to become more sustainable.
We are looking for an engineer to build the software and infrastructure that help scientists develop, evaluate, and deploy machine learning models for sustainability applications. You will work directly with applied scientists, turning production-grade science artifacts into reliable systems and building reusable tools that reduce the engineering work required for each new model.
The work spans research workflows and production systems: preparing data for experiments, automating training and evaluation, and making models available through dependable interfaces. A central design challenge is deciding what the platform should handle consistently and where researchers need flexibility. Good abstractions let scientists change a model or transformation without rebuilding the surrounding infrastructure, while keeping the underlying system accessible when something needs to be understood or fixed.
You will also develop interfaces and workflows that support AI-assisted engineering. This includes making platform capabilities understandable to coding agents and building checks that verify their changes before those changes reach production.
Key job responsibilities
- Build reusable data and ML infrastructure. Develop libraries, configuration-driven interfaces, and infrastructure-as-code components for data processing, model training, and inference.
- Build tools for repeatable experiments and model comparisons, preserving the data, code, and configuration needed to investigate results.
- Develop and evaluate AI-assisted workflows. Create tooling that helps coding agents use the platform correctly. Assess their output through automated checks and appropriate review, and measure whether the workflows improve engineering productivity without weakening reliability.
- Support adoption and evolution. Work with platform users to identify recurring needs, document design trade-offs, and maintain compatibility as shared components change.
About the team
Our team brings together applied scientists and engineers to develop models and software for sustainability applications. We work across research and production, building both scientific methods and the systems needed to use them reliably. That combination keeps platform development close to the practical needs of researchers and the people who depend on their models.
We build reusable tools so that each new research project does not require its own infrastructure. The engineering challenge is to make experimentation easier while preserving the checks and evidence needed to understand results and operate models in production. Scientists and engineers work together on those decisions, including where a shared approach helps and where a scientific problem requires something different.