Distributed Systems EngineerTurn billions of advertising signals into real-world outcomes — building the distributed systems that power ML model execution, audience activation, and campaign optimization for the world's largest brands on Amazon Marketing Cloud (AMC). You'll own services that process millions of audience memberships daily and orchestrate customer ML workloads with strict reliability, privacy, and latency requirements.The AMC Custom Models and Signals team owns two high-impact domains: our Custom Models platform (ML orchestration via AWS Clean Rooms, SageMaker, and Spark) and our audience and custom signals publication pipeline (high-throughput identity resolution and segment delivery to Amazon DSP and Sponsored Ads). You'll work directly with large enterprise advertisers and see your work drive measurable advertising outcomes.Key job responsibilities:Design, build, and operate distributed services (Java, Lambda, Step Functions, DynamoDB, Spark) that orchestrate ML model training, inference, and audience activationOwn end-to-end delivery of features across the Custom Models platform and audience/signals publishing pipeline — from API design through deployment and production monitoringInvestigate and resolve complex production issues spanning multiple services and AWS dependencies (Step Functions, Clean Rooms, SageMaker, Glue, S3)Improve system reliability, observability, and operational tooling for multiple Ads Tier 1 servicesCollaborate with science, product, and partner teams to translate customer requirements into scalable technical solutionsParticipate in on-call rotation and drive operational excellence through COEs, runbooks, and automated monitoringAbout the team:AMC Custom Models and Signals is a Boulder-based team building the systems that turn AMC analytics into real-world advertising outcomes.
We own Custom Models (ML orchestration), Igno (audience publishing), AMS (identity resolution and delivery), and Custom Signals API (new activation signals). Our work directly powers audience targeting, measurement, and ML-driven optimization for enterprise advertisers. We value ownership, operational rigor, and solving hard problems simply — and we're investing heavily in GenAI and agentic approaches to platform reliability and customer experience.