Software Development Engineer, Devices Software & Services

The Amazon Devices Software & Services team is looking for a Software Development Engineer to join a high-velocity, specialized team. This is not a traditional framework or infrastructure role - we are building a new groundbreaking customer experience that will define how millions of users worldwide interact with their technology - making it more proactive, intuitive, and invisible.

We’re building the systems that put AI into production — the runtime, services, and tooling that turn large language models and autonomous agents into reliable, safe, scalable products. As a Software Development Engineer II, you’ll design and ship core software in this stack: model and tool integrations, the agent execution loop, evaluation and quality pipelines, and the APIs that product teams build on. You’ll own meaningful components end to end, make real design decisions, and write production code that runs against real users and real data.
This is a hands-on building role for an engineer who wants to go deep on AI infrastructure — not just call an API, but engineer the reliability, safety, and performance around it. You’ll also work in an AI-native development environment: we use AI coding assistants and agentic workflows as part of our daily software-development lifecycle, and we hold that velocity to a hard bar of automated testing and code review. We expect engineers who can move fast and keep the quality gates green.

Key job responsibilities

• Invent New Paradigms: Move beyond traditional voice UI to build "Ambient Intelligence" experiences that span Alexa, Fire TV, and mobile/wearable surfaces, focusing on proactive assistance rather than reactive commands.
• Rapid Innovation: Operate in a fast-paced environment where you will prototype, test, and deploy features that leverage multimodal inputs (camera, audio, and documents) to create personalized customer magic.
• Cross-Surface Continuity: Ensure that a customer’s experience is persistent and context-aware beyond the device boundaries - whether they are at their desk using Alexa.com, in their living room with Fire TV, or on the go with a wearable.
• Design, build, test, and operate scalable backend services and libraries that power AI/LLM-driven features.
• Integrate and harden model providers, tools, and agent capabilities behind clean, versioned interfaces — with sane failure behavior, observability, and cost/latency awareness.
• Build evaluation and regression tooling so AI behavior is measured, gated, and trustworthy — not vibes.
• Apply security-conscious, fail-closed design to anything touching credentials, user data, tool execution, or autonomous actions.
• Work fluently alongside AI coding assistants: decompose tasks an agent can execute, critically review AI-generated code, and keep engineering rigor high at AI speed.

A day in the life
In our team, you will get to build new products and features from the ground up. You will define and build features and tools used by thousands of developers for building for millions of customers. We work across multiple internal teams as well as product to define what we are building and make it a reality. SDEs on this team are actively writing code, as well as working on designs and mentoring others. As the Device OS org we get to work horizontally across many exciting new device types.

About the team
We’re a platform team that builds the shared AI/agent infrastructure used across multiple products. We care about correctness, safety, and developer experience in equal measure. We operate test-first, review adversarially, keep our core clean and reusable, and treat AI as a tool that raises the engineering bar.