SoftwareDevelopmentEngineer—SPS,AgenticAIExperiences,SellingPartnerDevelopmentServices

The Spectrum team under Selling Partner Services umbrella owns the Selling Partner API (SP-API) platform — the authorization, identity, and Solution Provider experience layer that multiple third-party Solution Providers depend on to build applications powering millions of Amazon sellers. Our services process billions of API calls, manage complex permission models across global marketplaces, and are the front door for every Solution Provider building on Amazon's commerce platform. We're looking for an experienced SDE to lead the design and delivery of agentic AI experiences that transform how Solution Providers interact with our platform. This isn't bolting a chatbot onto an existing product — it's rethinking Solution Provider workflows from first principles using AI agents that can reason, act, and learn. You'll define the architecture for how AI agents interact with Spectrum's authorization, notification, and identity systems, and you'll ship production features that measurably reduce Solution Provider friction at scale. Key job responsibilities * Own the architecture of agentic AI systems that operate across Spectrum's service mesh — agents that can query authorization systems, assess qualified permissions, inspect Control Plane authorization decisions, and synthesize solution provider-facing guidance autonomously * Leverage AI-DLC to build world class experiences for Sellers and Solution Providers to discover and leverage agentic capabilities in the choice of their interface including Seller central, Amazon Quick, Claude, etc. * Design and build AI agent frameworks including tool-use orchestration, context retrieval (RAG over developer docs, API specs, and account data), guardrails, and evaluation pipelines * Lead end-to-end delivery of features from ambiguous problem statements through design, implementation, and production launch — writing LLDs, driving cross-team alignment, and mentoring junior engineers * Build intelligent automation for high-volume operation