Azure AI Search Overview
Azure AI Search is a fully managed, cloud-hosted service that connects your data to AI. The service unifies access to enterprise and web content so agents and LLMs can use context, chat history, and multi-source signals to produce reliable, grounded answers.What is Azure AI Search?
Azure AI Search is an AI-powered information retrieval platform that helps developers build rich search experiences and generative AI apps that combine large language models (LLMs) with enterprise or web data. Common use cases include classic search and modern retrieval-augmented generation (RAG) via agentic retrieval. This makes Azure AI Search suitable for both enterprise and consumer scenarios, whether you’re adding search functionality to a website, app, agent, or chatbot.Key Capabilities
When you create a search service, you unlock the following capabilities:Two Search Engines
- Classic search: Single requests for predictable, low-latency queries
- Agentic retrieval: Parallel, iterative, LLM-assisted search for complex agent-to-agent workflows
Query Types
- Full-text search: Traditional keyword-based search with BM25 relevance ranking
- Vector search: Semantic similarity search using embeddings
- Hybrid search: Combined full-text and vector search for optimal results
- Multimodal search: Query across text and images in a single pipeline
AI Enrichment
- Chunk, vectorize, and transform raw content to make it searchable
- Built-in skills for OCR, entity recognition, key phrase extraction, and more
- Integrated vectorization with Azure OpenAI and Foundry Tools
Enterprise Features
- Security: Azure scale, security, monitoring, and compliance
- Access control: Document-level permissions and role-based access
- Integrations: Azure OpenAI, Microsoft Foundry, Azure data platforms
Why Use Azure AI Search?
Ground AI Agents
Provide agents and chatbots with proprietary, enterprise, or web data for accurate, context-aware responses
Multi-Source Data
Access data from Azure Blob Storage, Azure Cosmos DB, SharePoint, OneLake, and other supported data sources
Hybrid Search
Combine full-text search with vector search to balance precision and recall
Production Ready
Enterprise security, access control, and compliance through Microsoft Entra and Azure Private Link
Classic Search vs Agentic Retrieval
| Aspect | Classic Search | Agentic Retrieval |
|---|---|---|
| Search corpus | Search index | Knowledge source |
| Search target | One index defined by schema | Knowledge base with multiple sources |
| Query plan | No plan, just a request | LLM-assisted or user-provided plan |
| Query request | Search documents in an index | Retrieve from knowledge sources |
| Response | Flattened search results | LLM-formulated answer, activity log, references |
| Status | Generally available | Public preview |
Getting Started
Create a Search Service
Deploy an Azure AI Search service in your preferred region and choose a pricing tier
Choose Your Path
Before getting started, make these key decisions:Choose a Search Engine
- Classic search: Best for traditional app needs with lower costs and complexity
- Agentic retrieval: Ideal for agent workflows and complex RAG scenarios requiring LLM orchestration
Choose a Region
- Select a region that supports your required features
- For agentic retrieval, choose a supported region
- Consider proximity to your data sources and users
Choose an Ingestion Method
- Pull method: Use indexers for supported data sources (Azure Blob, Cosmos DB, SQL, etc.)
- Push method: Upload JSON documents directly for real-time sync or unsupported sources
Do You Need Vectors?
- Required for semantic similarity search and embedding-based retrieval
- Use integrated vectorization to automate embedding generation
- LLMs and agents don’t require vectors unless using similarity search
Architecture
Classic Search Architecture
Agentic Retrieval Architecture
Next Steps
Create a Service
Deploy your Azure AI Search service
Quickstart
Build your first search index
Vector Search
Learn about vector search concepts
Agentic Retrieval
Explore agentic retrieval for AI agents