Description

Advertisers and advertising tools are rapidly adopting AI agents to plan, optimize, and achieve marketing goals. The Ads Agent Fabric team’s mission is to deliver a common infrastructure supporting all advertising-aware model-based applications (i.e., agents) and accelerate development of agentic applications for our advertisers. The team is responsible for an agent-builder SDK with pre-integrated Ads features, agent-builder tools that accelerate develop, test, and model evaluation lifecycle, and a fully-managed runtime with integrated access to the Ads services, data, and foundational models.

As the SDM on the Agent Fabric team, you will lead a team of engineers in the design and delivery of critical infrastructure components that enable Amazon Ads to build and deploy AI agents at scale. You'll work at the intersection of distributed systems, machine learning infrastructure, and developer tooling, defining technical strategy for how thousands of advertisers interact with AI-powered advertising solutions. This role requires deep technical expertise in building scalable platforms and comfort with ambiguity as you'll be building in an emerging space where best practices and technical requirements are still be defined. You will need strong system design skills, ability to hire and develop a team of engineers, define technical strategy, and capability to influence senior stakeholders on architecture decisions and roadmap planning.

Key job responsibilities
- Own the full software development lifecycle from design to production launch of core infrastructure components and services
- Manage a team of software engineers from junior to senior levels
- Build highly scalable, low-latency infrastructure supporting real-time agent execution
- Define technical standards and best practices for a team of engineers across development, testing, and deployment
- Partner with applied scientists to translate research innovations into production-ready services
- Lead architecture reviews and provide technical guidance to engineers across multiple teams