AI Deep Dive Series (Virtual) - Agent Memory for Building Robust, Accurate Agents скачать в хорошем качестве

AI Deep Dive Series (Virtual) - Agent Memory for Building Robust, Accurate Agents

Tech Talk: RAG is Dead. Long Live Agent Memory Speaker: Chris Latimer (Vectorize) Abstract: RAG solved a problem: how to ground LLMs in in external knowledge. This was useful for basic chatbots. But in a world where we want autonomous agents to solve business problems, RAG fundamentally doesn't fit. Agent memory systems are emerging as the new standard in AI architectures to solve problems like inconsistency, context rot, and hallucinations. In this talk, we'll identify the root causes of these problems and explain how agent memory allows you to build robust, accurate agents that learn over time. We'll do a deep dive into Hindsight, an open source agent memory system that achieves state of the accuracy on benchmarks and in production. We'll see how Hindsight uses biomimetic data structures to produce agent memory that works more like human memory - and how you can use these capabilities to deliver agents you can trust in production.