Software Development Engineer III, Prime Video Personalization & Discovery

Amazon; New York (NY); On-site; USD 180,000 - 260,000; Full time

Job summary

Amazon in New York invites a Senior Software Development Engineer to help build Prime Video's AI-powered conversational discovery experience. You will own end-to-end design, development, and operations for core platform components, collaborating with scientists and engineers to push the boundaries of how customers find content.

You will prototype and productionize LLM-powered features, optimize latency and cost, and contribute code with strong reliability and safety practices.

Responsibilities

• Own the design, implementation, and operation of core components of Prime Video's conversational discovery platform, delivering production-quality systems in a fast-moving, ambiguous environment
• Design hybrid approaches that combine LLM-based semantic understanding with traditional personalization and ranking signals for content discovery and recommendation
• Partner closely with applied scientists and fellow senior engineers to prototype, evaluate, and productionize LLM-powered features including personalized conversation starters, natural language query understanding, multi-turn dialogue, and contextual recommendations
• Optimize LLM inference for latency, cost, and quality at the scale of one of the world's largest streaming services
• Build evaluation and quality-gating mechanisms that measure relevance, personalization, diversity, and safety of LLM-generated content before and after launch
• Contribute significant code, conduct thorough code reviews, and raise the engineering bar across the team
• Uphold operational excellence for Tier-1 customer-facing AI systems, including on-call participation, root cause analysis, and driving reliability improvements
• Collaborate across team and organizational boundaries to integrate with partner systems spanning inference infrastructure, feature storage, and experimentation platforms

Job description

Description

Prime Video Personalization and Discovery (PVPD) is seeking a Senior Software Development Engineer to join a small, high-caliber team building the next generation of Prime Video's AI-powered content discovery experience. We are reinventing how customers find something to watch: instead of browsing and scrolling, customers tell Prime Video what they're in the mood for, through natural language conversation, personalized prompts, and interactive recommendations woven directly into the product. You'll work alongside a hand-picked group of senior engineers and applied scientists, each bringing deep expertise in their domain, moving fast on a high-visibility initiative with significant ambiguity. You'll own your area of the system end-to-end: designing, building, and operating the services that combine large language models with Prime Video's personalization, catalog, and engagement signals to deliver conversational discovery at massive scale.

Key job responsibilities

• Own the design, implementation, and operation of core components of Prime Video's conversational discovery platform, delivering production-quality systems in a fast-moving, ambiguous environment
• Design hybrid approaches that combine LLM-based semantic understanding with traditional personalization and ranking signals for content discovery and recommendation
• Partner closely with applied scientists and fellow senior engineers to prototype, evaluate, and productionize LLM-powered features including personalized conversation starters, natural language query understanding, multi-turn dialogue, and contextual recommendations
• Optimize LLM inference for latency, cost, and quality at the scale of one of the world's largest streaming services
• Build evaluation and quality-gating mechanisms that measure relevance, personalization, diversity, and safety of LLM-generated content before and after launch
• Contribute significant code, conduct thorough code reviews, and raise the engineering bar across the team
• Uphold operational excellence for Tier-1 customer-facing AI systems, including on-call participation, root cause analysis, and driving reliability improvements
• Collaborate across team and organizational boundaries to integrate with partner systems spanning inference infrastructure, feature storage, and experimentation platforms