Join this hands-on session to explore how to power production-ready Azure AI Agent Services using Azure AI Foundry Knowledge with Retrieval-Augmented Generation (RAG). We’ll guide you through building a comprehensive knowledge base by combining unstructured data from Fabric OneLake Files using Azure AI Search, structured data via Fabric Data Agents, and real-time signals through Bing Web Grounding.

Attendees will learn how to optimize Azure AI Search’s vector indices to support scalable RAG architectures and ensure high retrieval precision. The session will cover evaluation techniques for both retrieval quality and agent output, with practical examples on how to embed observability and telemetry into your solution. We’ll also demonstrate how the Model Context Protocol accelerates agent development cycles, making it easier to iterate and deploy AI solutions quickly.

Whether you’re building copilots, assistants, or domain-specific agents, this session provides the technical foundation and real-world patterns you need to get started.