Software Development Engineer, Benchmarks

Build the AI-agent-powered data platforms that process billions of events daily to help advertisers understand what "good" looks like — sitting at the intersection of petabyte-scale data engineering, real-time distributed systems, and innovative AI agent interfaces. This isn't a team that talks about AI — we're actively shipping it into production.

We're building agent-friendly APIs that let AI systems autonomously query, interpret, and surface benchmark insights to advertisers in real time. We're deploying operational AI agents that cut incident resolution times in half, and using GenAI-driven development practices to ship faster. On the data side, our infrastructure exceeds petabyte scale with millions of transactions per second — we recently migrated our data lake to Apache Iceberg, built real-time streaming pipelines with Spark, and designed unified reporting solutions that serve both human UX and AI agent consumption patterns. Our work regularly features at Amazon Ads' unBoxed flagship conference.

What makes this role stand out:

• You'll build systems consumed by both humans and autonomous AI agents — a new paradigm in advertising data
• Petabyte-scale infrastructure with millions of transactions per second
• GenAI-accelerated development practices — from spec-driven development to autonomous testing tools
• Stable, well-built systems with reasonable on-call focused on continuous improvement, not firefighting

Key job responsibilities

• Design and build scalable, real-time data systems that power advertiser-facing measurement products and AI agent interfaces
• Own end-to-end delivery: requirements, design, implementation, testing, deployment, and operational support
• Build monitoring, alerting, and data quality frameworks that ensure reliability at petabyte scale
• Optimize distributed systems for performance, cost efficiency, and extensibility
• Contribute to our GenAI-accelerated development practices — from spec-driven development to autonomous testing tools
• Collaborate with scientists, product managers, and partner teams to translate complex measurement problems into elegant engineering solutions
• Navigate ambiguity and make sound technical decisions with incomplete information

About the team

We're part of Measurement and Data Science (MADS) — the scientific and engineering center of excellence for campaign measurement across Amazon Advertising. Our charter is Benchmarking and Brand Metrics: we compute peer comparison insights across millions of brands worldwide and deliver them through real-time APIs, batch pipelines, and AI agent integrations. We're collaborative, technically ambitious, and value sustainable engineering.