Software Development Engineer, AWS Payments & Fraud Prevention

What if the data you engineer could stop fraud before it happens - and protect millions of AWS customers worldwide?
The Payments and Fraud Data Platform team within AWS Payments & Fraud is the backbone of one of the most critical missions at AWS: ensuring every transaction is safe, every suspension is justified, and every customer experience is seamless. We process petabytes of data daily, powering the ML models, rule engines, and real-time reporting systems that allow us to detect bad actors.
We are looking for a Software Development Engineer who is passionate about building large-scale data systems, thrives in ambiguity, and wants to work at the intersection of big data, machine learning, non-payment and fraud prevention.

Key job responsibilities
Architect & Build Petabyte-Scale Data Pipelines - Design, develop, and optimize batch and near-real-time data pipelines using Amazon Managed Airflow (MWAA), Fargate, EMR, Apache Iceberg, and Amazon Redshift to ingest, transform, and serve data at massive scale.
Power ML & Rules Engines - Deliver high-quality, timely datasets that fuel non payment and fraud detection ML models and business rule engines, directly impacting fraud prevention and non-payment suspension accuracy.
Build Intelligent Data Access Layers - Design and implement a Smart Query & Data Access Analyzer that profiles query patterns, identifies hot data, and automatically promotes frequently accessed datasets into hot memory tiers (e.g., OpenSearch, DynamoDB DAX, Redshift RA3 caching) for sub-second retrieval.
Create Natural Language SQL Interfaces - Leverage Amazon Bedrock and LLM (large language models) to build a natural language-to-SQL generation tool that empowers analysts and stakeholders to query complex fraud datasets conversationally - no SQL expertise required.
Develop Real-Time Dashboards & Insights - Build rich, interactive dashboards using Grafana and internal tooling that provide real-time visibility into fraud trends, pipeline health, data quality metrics, and operational KPIs.
Curate a Living Data Dictionary - Design and maintain a comprehensive, searchable data catalog and dictionary that provides rich metadata, lineage, and documentation - improving discoverability and self-service analytics across the organization.
Champion Data Quality at Scale - Implement automated data validation, anomaly detection, and reconciliation frameworks that ensure accuracy, completeness, and consistency across the entire data lifecycle.
Secure & Govern Sensitive Data - Implement encryption, fine-grained access controls, audit trails, and compliance frameworks to protect highly sensitive payments and fraud data.
Monitor, Optimize, Repeat - Build observability into everything , proactive alerting, pipeline performance tracking, SLA monitoring, and intelligent auto-remediation to ensure five-nines data reliability.

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, batch and near real-time systems that ingest billions of signals daily, process petabytes of data, and make timed decisions to stop bad actors from account registration through the entire customer lifecycle. Our platform is built on a modern, cloud-native architecture designed for petabyte-class batch processing. Our stack includes Amazon MWAA orchestrates workflows, while EMR and Fargate handle scalable data processing. Amazon Redshift powers our data warehouse, complemented by an Apache Iceberg lakehouse for open table format flexibility. DynamoDB and OpenSearch drives search and analytics, and Amazon Bedrock serves as our AI/ML foundation. Development is accelerated with Kiro, an AI-powered IDE, and Grafana provides full-stack observability.
We are a team of builders who care deeply about customer trust and love solving complex problems at massive scale.