As part of AI Operations Integration, we have a vision to transform Amazon Operations & Supply Chain into an AI-Native organization by delivering intuitive and differentiated AI solutions that solve enduring operational challenges. We blend vision with curiosity and Amazon's real-world experience to build rapidly AI capabilities. We accelerate our customers' businesses: internal operations teams across Amazon's global footprint through delivery of predictive analytics, LLMs and autonomous AI agents to automate decision-making across Amazon's global operations and supply chain.

As a Machine Learning Software Development Engineer, you'll design and deploy production ready systems that combine traditional ML with modern agentic architectures to solve impactful operational problems at scale.

This role combines the excitement of a startup environment with the scale of Amazon Operations. You'll research state-of-the-art open source and internal tools and will tackle highly ambiguous problems. If you thrive on ownership and dealing with ambiguity, passionate about AI and want to fundamentally influence how Amazon Operations leverages AI, this role offers an extraordinary opportunity to make your mark.

Key job responsibilities
What you'll do:

• Build and deploy ML/LLM-powered features across the full stack - from data pipelines and model serving to user-facing internal tools
• Implement AI agent components that automate complex operational workflows across multiple systems and decision points
• Develop internal front-end applications (dashboards, tools, products) that make AI outputs accessible to non-technical operations users at scale
• Build integrations across internal APIs, databases, and MCP servers to enable multi-system orchestration
• Contribute to the ML lifecycle: data pipelines, experimentation, deployment, monitoring, and evaluation
• Implement guardrails, evaluation frameworks, and human-in-the-loop patterns for production AI systems
• Collaborate with operations teams, program/product and scientists to translate requirements into shipped software

The ideal candidates are engineers who've shipped ML/LLM systems to production, understand the practical challenges of MLOps, and want to work on problems where the impact is measurable in both customer experience and operational efficiency.

About the team
We're a cross-functional team of machine learning engineers, ML & AI scientists and Technical PM focused on automating operational decisions in Amazon's supply chain.

Our charter is to identify high-impact automation opportunities, build AI agents that can handle them reliably, and deploy these systems into production
where they process real decisions daily.

The team operates with a build-measure-learn cycle. We work closely with operations partners to understand their problems, prototype solutions quickly, measure impact rigorously, and iterate based on real-world performance.