Amazon Advertising sits at the intersection of e-commerce and advertising, helping brands reach customers across owned surfaces, the open web, and streaming media. We build with the advertiser as the starting point and work backwards from what they need to run effective, measurable campaigns at scale.
We're looking for a Sr. Software Development Engineer to join the Personalization & Guidance team — a high-growth team building AI-native systems that make advertising more intelligent, transparent, and effective for the advertisers who rely on them every day.
Advertising personalization is a well-studied domain. What makes this team's work different is the explainability constraint — we're building systems where the why behind a recommendation is a first-class product concern, not an afterthought. That drives real architectural decisions: how you store and version the inputs to a decision, how you surface confidence, how you maintain auditability across distributed agent systems, and how you enforce content quality standards at render time rather than at authoring time.
Key job responsibilities
You'll work on systems that sit at the core of how advertisers understand and act on their advertising performance. The problems are technically hard:
- Personalization at scale — building and maintaining structured models of advertiser behavior, goals, and preferences that update continuously across millions of accounts and inform real-time decisions
- Real-time ML inference — systems that evaluate and rank recommendations, guidance, and actions at high volume with tight latency constraints across distributed infrastructure
- Explainable AI systems — infrastructure that makes automated decisions auditable and legible to end users, not just to engineers. This includes stable data models for tracking decision inputs, surfacing confidence signals, and rendering explanations that advertisers can inspect and act on
- Multi-agent coordination — APIs and contracts that allow multiple specialized AI agents to contribute outputs into a unified experience, with quality enforcement at the delivery layer
- Event-driven data pipelines — high-volume ingestion, processing, and serving of advertising signals that feed both the ML models and the real-time recommendation systems
A day in the life
You'll work in a space where ML infrastructure, distributed systems, and product-quality UX concerns intersect daily.
Technology You'll Work With:
- Large-scale distributed systems built on cloud infrastructure (microservices, event streaming, real-time serving)
- Machine learning infrastructure — model serving, feature stores, A/B testing frameworks, online evaluation
- Large language models and agentic AI systems — integrating LLM-powered agents into production advertising workflows, with the reliability and observability requirements that entails
- High-volume data pipelines processing advertising signals across formats and surfaces
- API design for multi-team and multi-agent consumption patterns
About the team
What We're Looking For:
- Strong distributed systems fundamentals — you've built services that have to be both fast and correct under real traffic
- Experience with ML infrastructure or production ML-adjacent systems; you understand what it takes to build reliable systems around probabilistic models
- Product instincts — you care that the output is useful and legible to the person using it, not just that the pipeline ran successfully
- Comfort with ambiguity — the team is working in a fast-moving space where some of the right architectural decisions are still being established
- Strong written communication — our most consequential technical decisions live in design documents, and engineers are expected to author them
If you want to work on systems that are technically sophisticated, have real daily impact on advertisers around the world, and are building into a space — AI-driven advertising — that is genuinely changing how the industry works, this team is worth a conversation.
