Agentic Search in Practice: Building Agentic Retrieval Systems with Amazon OpenSearch
Modern AI search applications are becoming increasingly complex—combining keyword, semantic, and vector retrieval with LLM-powered reasoning. But the next leap goes further: systems that can think, iterate, and act like agents.
In this session, we’ll show developers how to build agentic search applications on Amazon OpenSearch—systems that understand user intent, dynamically refine queries, and execute multi-step workflows to deliver meaningful outcomes with latest innovations from Amazon OpenSearch.
We’ll walk through practical architectures and patterns, including:
· Hybrid and vector search pipelines
· Agent-driven query planning and execution
· Building agentic search experiences within modern agentic IDEs using Agent Skills
Whether you're developing RAG applications or next-generation AI assistants, this webinar will equip you with the tools and design patterns to move beyond retrieval—toward intelligent, action-oriented Agentic search systems powered by Amazon OpenSearch.
Speaker Bios:
Bobby Mohammed
Bobby Mohammed is a Principal Product Manager at AWS, leading product initiatives at the intersection of Search, Generative AI, and Agentic AI—shaping how modern systems retrieve, reason, and act. His work focuses on next-generation intelligent applications, from retrieval-augmented generation (RAG) to agent-driven workflows and long-term memory systems. Previously, he helped build foundational AI and data capabilities on Amazon SageMaker, spanning data, analytics, and machine learning at scale. Prior to AWS, he served as Director of Product at Intel, leading deep learning training and inference platforms powering high-performance AI infrastructure. Bobby holds an MBA from the Kellogg School of Management at Northwestern University, and Master’s and Bachelor’s degrees in Electrical Engineering.
Sean Zheng
Sean Zheng is a Senior Engineering Manager at AWS, where he leads ML/GenAI and search relevancy components within AWS OpenSearch. His team owns plugins including ML Commons, Neural Search, and Search Relevancy Workbench, serving as the primary driver of ML and agentic capabilities for OpenSearch. Recent deliveries under his team include Agentic Search, Agentic Memory, and a Python-based agentic service. Prior to his role with AWS OpenSearch, Sean worked across multiple teams in Amazon's retail organization, focusing on machine learning and data analytics. His experience spans Core ML, Product Graph, and Search Engine Optimization teams.Sean holds a PhD degree in Computer Science from State University of New York.