Amazon Ads Traffic ingests hundreds of billions of ad events daily, decorates them with dimensional and identity data, and publishes the structured traffic and billing datasets that the entire Ads measurement ecosystem depends on — including three of the five critical metrics used for Ads billing. You'll design and deliver features across Spark-based EMR pipelines, decoration jobs, and publication systems that operate under strict tier-one SLAs for data freshness and accuracy.
The Traffic Applications team, part of Measurement, AdTech, and Data Science (MADS), is the first system in the MADS stack to encounter regulatory and identity changes — meaning the architectural decisions made here propagate across the entire measurement pipeline. You'll work on high-visibility initiatives that directly impact Ads VP and SVP-level goals, including compliance work (GDPR, ePrivacy), geo-targeting features, and architecture scalability to support spiky traffic from live sports events.
Key job responsibilities
- Design and develop features across Spark-based EMR pipelines, decoration jobs, and publication systems that process hundreds of billions of ad events daily
- Deliver end to end across the full development lifecycle — from requirements and system design through implementation, testing, deployment, and operations under tier-one SLAs
- Navigate technical problems with competing constraints — regulatory deadlines, performance requirements, cost targets, and cross-org dependencies — and propose well-reasoned solutions
- Collaborate with engineers across Ads, Customer Trust, and identity-owning upstream systems to ensure Traffic's design decisions support downstream consumers
- Contribute to compliance initiatives (GDPR, ePrivacy) and scalability efforts for high-traffic events like live sports
- Maintain sound operational practices — alarms, telemetry, runbooks — for a system where data freshness and accuracy directly impact billing
- Mentor peers through design reviews, code reviews, and knowledge sharing
A day in the life
On a typical day you might start with a stand-up to align on priorities, then dive into building or extending a Spark pipeline to handle a new decoration requirement. After lunch, you could be debugging a data freshness issue against a tier-one SLA, tracing it through the ingestion path to a root cause. Between coding, you'll review pull requests, participate in design discussions for upcoming compliance or scalability work, and collaborate with downstream consumers to understand how they use your datasets.
About the team
The Traffic Applications team owns the data ingress system for all of Amazon Ads — ingesting hundreds of billions of ad events daily, decorating them with dimensional and identity data, and publishing structured traffic and billing datasets to consumers across the Ads ecosystem. The system operates under strict SLAs for data freshness and accuracy on tier-one advertising metrics, owning three of the five metrics used for billing. As the first MADS system to encounter regulatory and identity changes, Traffic's architectural decisions propagate across the entire MADS stack, making design choices here uniquely consequential.
