We have all been there—spending weeks fine-tuning indexes in Sitecore Search only to have users complain that they can't find what they need because they didn't use the exact "keyword." Traditional keyword matching is hitting its ceiling. While "AI Agentic Search" is the current buzzword, the reality of implementing it—handling content chunking, managing vector embeddings, and maintaining real-time sync—is where most projects struggle.

In this session, we explore how to build intelligent, AI-driven Agentic search solutions powered by Sitecore Search.

The Sitecore Search Ingestion Pipeline: How to use Experience Edge webhooks and the Sitecore Search Push API to handle near real-time content updates.

The Chunking Strategy: The extracted HTML is segmented into meaningful chunks that work for vector indexing.

Query Planning from User input – The user query, goes to LLM, where it is analysed and built into multiple enriched semantic queries.

Returning the Merged Results: These semantic queries are then executed against Vector Index and a relevant set of results from different queries are aggregated and reranked against original user intent using semantic similarity scoring.

Taking it a step further, we will be looking at building a RAG with Agentic Search in Sitecore, to enable context aware & a truly conversational search experience driven by LLM-based query orchestration.

By combining content chunking, vector indexing & intelligent query orchestration, this session demonstrates how Agentic Search can significantly improve search accuracy, relevance & user experience — transforming search from a keyword-based mechanism into an intelligent discovery engine.

What you will Learn:
1. How to use LLMs to think about Query planning.
2. How to use Semantic Search API of Sitecore Search & build a meaningful conversational experience.
3. How to architect Agentic Search using Sitecore AI & Sitecore Search.