Software Development Engineer, AWS Fraud Prevention

Job description

Are you passionate about building highly scalable distributed systems that protect millions of customers? Do you want to work on a team that has AI first approach to all problems? The AWS Payments & Fraud Prevention team is looking for a talented SDE to help us stay one step ahead of bad actors.

You'll design and build systems that process millions of events per second, develop machine learning-powered detection pipelines, and directly contribute to keeping AWS safe. Whether it's architecting a new real-time detection engine, optimizing our graph-based fraud network analysis, or hardening our registration fraud defenses your work will have immediate, measurable impact at AWS scale.

Key job responsibilities

- Design, build, and operate highly scalable distributed systems to create backend services and serverless workflows that take actions where correctness directly protects AWS customers and revenue.

- Own features end to end: write design docs, implement in Java/TypeScript on Lambda, DynamoDB, and EventBridge, and ship through CDK-managed pipelines that deploy to 38+ commercial regions plus GovCloud and sovereign clouds.

- Write high-quality, production-ready code for services that must be available 24/7 with zero tolerance for downtime.

- Own the full lifecycle of your solutions — from design and implementation through deployment, monitoring, and operational excellence.

- Build agentic and LLM-based automation that triages fraud signals and use AI developer tools daily.

- Learn from others, mentor teammates, and contribute to a culture of engineering excellence and continuous improvement.

- Collaborate with applied scientists to integrate machine learning models into real-time decisioning systems.

- Work in a fast-paced, agile environment where you're solving complex problems with direct, measurable impact on AWS security.

About the team

The AWS Payments & Fraud Prevention team protects AWS and its customers from fraud, abuse, and account compromise at global scale. We build highly scalable, real-time systems that ingest billions of signals daily, process petabytes of data, and make split-second decisions to stop bad actors — from account registration through the entire customer lifecycle. Our stack includes Apache Flink for stream processing, AWS Neptune for graph-based intelligence, and machine learning models for adaptive detection.

We're a team of builders who care deeply about customer trust and love solving complex problems at massive scale.