Learn to build superior AI agents using neural network-based Retrieval-Augmented Generation (RAG) in this 19-minute conference talk from the AI Engineer World's Fair. Watch Will Bryk, CEO of Exa.ai, live code two AI agent applications to demonstrate the practical differences between traditional keyword-based search RAG and neural network RAG via vector search. Compare both applications based on task performance, relevance, and latency through real-time demonstrations rather than theoretical explanations. Discover why traditional keyword-based search engines consistently underperform in agentic or multi-step tasks where semantic grounding and contextual nuance are crucial. Master embedding strategies, indexing trade-offs, hybrid retrieval techniques, and prompt tuning methods that improve RAG quality for AI agents in production environments.