Find your next role

Discover amazing opportunities across our network of companies committed to gender equality in the workplace.

Applied Science Manager - Web Search/Retrieval/Ranking, AGI Info - Web Information Systems

Amazon

Amazon

Sunnyvale, CA, USA
Posted on Oct 14, 2025

Description

Help shape the future of artificial intelligence by leading breakthrough innovations in web search and Retrieval Augmented Generation at Amazon AGI. We're seeking an exceptional Applied Science Manager to drive the development of next-generation search capabilities to improve agentic applications across Amazon. Your work will directly influence the next generation of Amazon's search and AI systems, improving how millions of users interact with our platforms through more intelligent and efficient information retrieval systems.

What will you do?: You will lead a team of scientists to improve our web search capabilities and drive key improvements across multiple agentic systems across Amazon. You will be responsible for: (i) developing novel retrieval and ranking models to improve search, incorporating multiple objectives (e.g. relevance and trustworthiness), and partnering closely with engineering to improve model performance; (ii) improve content and query understanding models to deliver improved signal to retrieval and ranking models; (iii) partner closely with content acquisition and customer teams to ensure our dependencies are met and we’re delivering value to the end customers, enhancing information grounding for LLMs; (iv) develop a science roadmap, including publication opportunities and how we can accelerate delivery of customer impact; (v) coach and develop the team, hire, and hold the bar on scientific rigor throughout the team.

Technical Focus Areas: An ideal candidate will have demonstrated experience in Information Retrieval Systems (learned sparse and dense retrieval, efficient sparse and ANN indexing, learning-to-rank models, content understanding/information extraction, etc.) and Large Language Models (Retrieval-augmented generation (RAG) architectures, post-training techniques, inference optimization techniques, etc.).