About this talk
This talk addresses the challenges faced by agents when performing complex tasks, highlighting that the issue often lies in the lack of proper context for advanced reasoning. The speaker focuses on the importance of context engineering for creating reliable agentic applications, discussing methods for modeling information with explicit relationships and designing retrieval patterns that follow knowledge pathways. Additionally, the session covers the development of memory systems that preserve the connections necessary for sophisticated reasoning, making it essential for teams looking to enhance their agent capabilities.
