Find your next role

Description

Amazon Advertising is one of Amazon's fastest growing and most profitable businesses, responsible for defining and delivering AI-powered solutions that transform how advertisers make strategic decisions. We deliver billions of ad impressions and process massive volumes of advertiser data every single day. You'll work with us to pioneer breakthrough approaches in how AI agents access and reason over real-time advertiser data at scale.

We are using generative AI and agentic systems to help advertising agents provide instant, strategic advice to millions of advertisers. You will need to invent new techniques for agent orchestration, context optimization, and code generation to ensure we're delivering accurate, trustworthy insights with minimal latency and token consumption. You'll create feedback loops to ensure our solutions are constantly evaluating themselves and improving.

We're building the Ads Real-Time Data Service to solve this. We operate at the intersection of real-time data engineering and AI agent infrastructure. Our mission is to provide every advertising agent with immediate access to advertiser context—active campaigns, budget allocation, performance trends, brand positioning, product portfolios—from the first prompt. We're building summarized data for context using a mix of state of the art techniques like CodeAct and RAG-based embeddings. Our job is a multi-dimensional optimization problem: we need to ingest data from our data warehouse and Kafka streams, process billions of data points with 1-3 minute refresh cadences, deliver via Model Context Protocol (MCP) servers with sub-second latency, and enable agents to use CodeAct patterns (code execution in isolated sandboxes) for analytical queries. We need to do this while serving 30+ advertising agents and skills with high availability.

Key job responsibilities
System Architecture & Design
- Design and implement scalable architectures for real-time data ingestion from our data warehouse and Kafka streams processing billions of data points daily
- Build Model Context Protocol (MCP) server infrastructure—an emerging standard for AI agent-data interaction—that delivers advertiser context with sub-second latency and minimal token consumption
- Architect data processing pipelines that generate pre-computed advertiser context using the latest AI patterns
- Establish system design patterns and best practices for real-time data delivery at scale

Real-Time Data Engineering
- Develop high-throughput data ingestion systems handling both batch and streaming data sources with 1-3 minute refresh cadences
- Implement data processing logic to transform raw advertiser data into formats optimized for large language models with rich metadata
- Build monitoring and observability systems to track data freshness, processing latency, and system health across the platform
- Optimize data storage and retrieval patterns for sub-second query performance

Performance & Scalability
- Optimize system performance to achieve near-perfect response success rates and significant token reduction versus traditional approaches
- Build distributed systems that scale horizontally to support advertising agents and skills with growing data volumes
- Implement caching strategies and query optimization techniques to minimize latency and computational costs
- Conduct load testing and capacity planning to ensure platform reliability under peak traffic

AI-Assisted Development & Operational Excellence
- Use AI coding agents like Kiro to generate technical specifications and implementation code, accelerating development from weeks to days
- Participate in rapid spec-to-code workflows where specifications are approved and implemented within a single day
- Participate in on-call rotation supporting production systems with 99.9%+ availability requirements
- Write comprehensive documentation, runbooks, and technical designs for platform components
- Collaborate with cross-functional teams to understand requirements and drive adoption
- Mentor junior engineers through code reviews, design discussions, and knowledge sharing

A day in the life
Morning: AI-Assisted Spec Generation and Review

You start your morning reviewing overnight metrics: 2.3 billion data points processed, 47-second average latency, zero errors. You notice a slight uptick in processing time for data warehouse ingestion and dive into the logs to investigate. You identify an opportunity to optimize the data transformation logic and use Kiro to generate a technical specification outlining the proposed changes, including performance benchmarks and rollback strategy. Within 30 minutes, you have a complete spec ready for review.

Mid-morning, you're in a design review discussing a new technical specification for the advertiser context framework—also generated using Kiro. A teammate has prompted Kiro to create a schema-driven approach that lets agent teams define their data requirements declaratively. You review the AI-generated spec, discuss trade-offs between flexibility and performance, and suggest refinements to the error handling section. The team agrees on the approach and you use Kiro to generate the initial implementation, which you'll refine and test this afternoon.

Afternoon: Rapid Implementation and Collaboration

After lunch, you're refining the MCP server enhancements that Kiro generated this morning based on your approved specification. You're adding support for newly embedded data, and running evaluations to ensure its meets the benchmark. You review Kiro's implementation of the security requirements, add additional test cases, and run the comprehensive test suite. Everything passes, and you submit the pull request—spec to production-ready code in a single day.

You join a meeting with an advertising agent team. They're asking about adding brand metrics data to the platform. You walk them through the onboarding process, estimate it'll take less than a day with the new advertiser context framework, and commit to supporting their launch timeline. After the meeting, you use Kiro to generate a specification for their requirements and add it to the backlog for tomorrow's implementation.

Late Afternoon: Code Review and Innovation

You're reviewing a pull request from a junior engineer who used Kiro to implement a feature from this morning's approved specification. The AI-generated code follows the spec well, and the engineer added thoughtful refinements. You suggest some improvements to error handling and approve the pull request. You watch the CI/CD pipeline deploy it to production and monitor the metrics to ensure everything looks healthy.

About the team
The Ads Real-Time Data Service team is a highly motivated, collaborative and fun-loving group of engineers building the foundational platform for Amazon's advertising AI future. We are entrepreneurial and have a bias for action with a broad mandate to experiment and innovate. Our team operates at the intersection of real-time data engineering, AI agent infrastructure, and distributed systems engineering—solving problems that directly impact how millions of advertisers interact with Amazon's advertising products.

We value technical excellence, customer obsession, and sustainable engineering practices. Our team includes engineers with diverse backgrounds in distributed systems, real-time data processing, AI/ML infrastructure, and platform engineering. We celebrate innovation (patent submissions encouraged), knowledge sharing (weekly tech talks), and continuous learning. We maintain a sustainable pace with minimal on-call burden, flexible work arrangements, and a strong focus on work-life balance. We're at the forefront of AI-assisted development, using tools like Kiro to accelerate our development cycles from weeks to days.