​R.I.P. Naive RAG - Welcome Agentic Retrieval

​Hosted by: HORNET.dev

​The era of single-turn, "chunk-and-embed" RAG is over. The future lies in agentic retrieval. From using LLMs as answer-generators into active search operators.

​If you've moved past the "RAG 101" tutorials and want to understand how to build more powerful retrieval for AI systems, this event is for you.

​Speakers: 

• ​Jo Kristian Bergum, CEO @ HORNET.dev 

• ​Zain Hasan, AI @ Together AI Together.ai

• ​Hamel Husain @ Parlance Labs

• ​Bryan Bischof AI @ Theory Ventures

​Location: https://www.shack15.com/

​18:00 Doors open & check-in 

​Grab a drink. Find the other builders.

​18:30 Autopsy of RAG: What we’re killing & what we're keeping 

​A brutal look at the failures of the chunk-and-embed model. We’ll dissect where it falls apart and what patterns are actually worth salvaging. We'll also speed-run through the taxonomy of the ‘R’ in RAG, from embeddings, through BM25, to graph-based methods.

​19:00 The Agentic Shift: LLMs are shockingly good at search 

​How agents turn retrieval from a single-query to an active multi-step retrieval process. We'll cover multi-hop querying, retrieval-as-a-tool, and strategies that make LLM-driven retrieval work.

​19:30 A No-BS panel

​When is a simple, battle-tested RAG pipeline good enough? When is the complexity of an agentic system justified? What about multimodal?, and is BM25 the answer to everything? We’ll open the mic.

​20:15 Drinks & debrief 

​More drinks. Light snacks. No slides. Corner the speakers or network with other builders.