From Search to Agents: Building RAG and Agentic Workflows

Want to build applications that can search intelligently and act on their own? In this hands-on workshop, we will walk through how to. Most AI demos fall apart at the retrieval step. The model is fine. The data is the problem. This session walks through how Elasticsearch fits into the AI stack, starting from plain search and building up to agents that can actually do things. We will cover what changes when you move from keyword search to semantic and hybrid retrieval, why RAG is mostly a retrieval engineering problem rather than a prompting one, and where agentic patterns genuinely help versus where they add cost and latency for no reason. We will also look at Elastic Agent Builder and how it turns your indices into tools an agent can call. You will leave knowing how to set up semantic search on your own data, how to structure a retrieval layer that a model can actually use, and what to try next on a free Elastic Cloud trial. No credit card needed, so you can follow along or build after. Bring a laptop if you want to try things live. Who should attend: Developers, data engineers, and anyone building search or AI features who wants the practical version, not the hype version. Want to talk? Have a 5-10 minute lightning talk or a deeper 45-minute session on how you built something? We’re looking for speakers. Submit through our CFP or email [email protected] Stay in the loop Join the Elastic Ahmedabad User Group to get updates on this and future meetups.