Azure AI Search

https://learn.microsoft.com/en-us/azure/search/
Admin
Azure AI Search is a cloud search service that enables information retrieval at scale with support...
Tokens:1,357,849
Snippets:8,830
Trust Score:10
Update:1 week ago
Show doc for...
Tokens:
Raw
### Install dependencies

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-powershell

Installs the npm dependencies for the quickstart project. After completion, a node_modules folder should appear.

```bash
npm install

```

--------------------------------

### Clone the sample repository

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-powershell

Clones the sample repository containing the quickstart notebook. Requires Git to be installed.

```bash
git clone https://github.com/Azure-Samples/azure-search-python-samples

```

--------------------------------

### Set up development environment with Git, Maven, and Azure CLI

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-agentic-retrieval

Clone the sample repository, navigate to the quickstart folder, configure environment by renaming sample.env to .env, install dependencies with Maven, and sign in to Azure for keyless authentication. Replace placeholder values in sample.env before renaming.

```bash
git clone https://github.com/Azure-Samples/azure-search-java-samples

```

```bash
cd azure-search-java-samples/quickstart-agentic-retrieval

```

```bash
mv sample.env .env

```

```bash
mvn clean dependency:copy-dependencies

```

```bash
az login

```

--------------------------------

### Knowledge agent response example (JSON)

Source: https://learn.microsoft.com/en-us/azure/search/agentic-retrieval-how-to-migrate

Example response from the GET request, showing the structure of a knowledge agent definition.

```JSON
{
  "name": "earth-at-night",
  "description": "A sample knowledge agent that retrieves from the earth-at-night knowledge source.",
  "retrievalInstructions": null,
  "requestLimits": null,
  "encryptionKey": null,
  "knowledgeSources": [
    {
      "name": "earth-at-night",
      "alwaysQuerySource": null,
      "includeReferences": null,
      "includeReferenceSourceData": null,
      "maxSubQueries": null,
      "rerankerThreshold": 2.5
    }
  ],
  "models": [
    {
      "kind": "azureOpenAI",
      "azureOpenAIParameters": {
        "resourceUri": "<redacted>",
        "deploymentId": "gpt-5-mini",
        "apiKey": "<redacted>",
        "modelName": "gpt-5-mini",
        "authIdentity": null
      }
    }
  ],
  "outputConfiguration": {
    "modality": "answerSynthesis",
    "answerInstructions": null,
    "attemptFastPath": false,
    "includeActivity": null
  }
}

```

--------------------------------

### Navigate to quickstart folder and open in Visual Studio Code

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-agentic-retrieval

Changes to the Quickstart-agentic-retrieval folder and opens it in Visual Studio Code. Run after cloning the sample repository.

```bash
cd azure-search-rest-samples/Quickstart-agentic-retrieval
code .

```

--------------------------------

### Install required Python packages for CMK encryption

Source: https://learn.microsoft.com/en-us/azure/search/search-security-manage-encryption-keys

Install the necessary Python packages: python-dotenv, azure-core, azure-search-documents==11.5.1, and azure-identity. These are required for the subsequent examples.

```Python
! pip install python-dotenv
! pip install azure-core
! pip install azure-search-documents==11.5.1
! pip install azure-identity

```

--------------------------------

### Build and run AgenticRetrievalQuickstart on Windows

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-agentic-retrieval

Compile and run the Java quickstart on Windows using a semicolon-separated classpath with the dependency wildcard. Run from the project directory after dependencies are downloaded.

```bash
javac AgenticRetrievalQuickstart.java -cp ".;target\dependency\*"
java -cp ".;target\dependency\*" AgenticRetrievalQuickstart

```

--------------------------------

### Navigate to the quickstart folder and open in Visual Studio Code

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-vector

Changes directory to the vector search quickstart folder and opens it in Visual Studio Code. Run after cloning the repository.

```bash
cd azure-search-python-samples/Quickstart-Vector-Search
code .

```

--------------------------------

### Get Indexer Status response example

Source: https://learn.microsoft.com/en-us/azure/search/search-monitor-indexers

Example response from the Get Indexer Status command. The top-level status indicates the indexer is set up correctly and available to run, not that it is currently running. Each run has its own status: running, success, transientFailure, or persistentFailure. A reset adds an entry with a Reset status.

```json
{
    "status":"running",
    "lastResult": {
        "status":"success",
        "errorMessage":null,
        "startTime":"2018-11-26T03:37:18.853Z",
        "endTime":"2018-11-26T03:37:19.012Z",
        "errors":[],
        "itemsProcessed":11,
        "itemsFailed":0,
        "initialTrackingState":null,
        "finalTrackingState":null
     },
    "executionHistory":[ {
        "status":"success",
         "errorMessage":null,
        "startTime":"2018-11-26T03:37:18.853Z",
        "endTime":"2018-11-26T03:37:19.012Z",
        "errors":[],
        "itemsProcessed":11,
        "itemsFailed":0,
        "initialTrackingState":null,
        "finalTrackingState":null
    }]
}
```

--------------------------------

### Get knowledge source definition with SearchIndexClient or HTTP

Source: https://learn.microsoft.com/en-us/azure/search/agentic-knowledge-source-how-to-fabric-ontology

Retrieves a single knowledge source by name to review its JSON definition. The .NET and Python examples use the SearchIndexClient SDK; the HTTP example uses the Knowledge Sources - Get operation. Requires a valid search endpoint and credentials.

```csharp
using Azure.Search.Documents.Indexes;
using System.Text.Json;

var indexClient = new SearchIndexClient(new Uri(searchEndpoint), credential);

// Specify the knowledge source name to retrieve
string ksNameToGet = "earth-knowledge-source";

// Get its definition
var knowledgeSourceResponse = await indexClient.GetKnowledgeSourceAsync(ksNameToGet);
var ks = knowledgeSourceResponse.Value;

// Serialize to JSON for display
var jsonOptions = new JsonSerializerOptions 
{ 
    WriteIndented = true,
    DefaultIgnoreCondition = System.Text.Json.Serialization.JsonIgnoreCondition.Never
};
Console.WriteLine(JsonSerializer.Serialize(ks, ks.GetType(), jsonOptions));

```

```python
# Get a knowledge source definition
from azure.core.credentials import AzureKeyCredential
from azure.search.documents.indexes import SearchIndexClient
import json

index_client = SearchIndexClient(endpoint = "search_url", credential = AzureKeyCredential("api_key"))

ks = index_client.get_knowledge_source("knowledge_source_name")
print(json.dumps(ks.as_dict(), indent = 2))

```

```http
### Get a knowledge source definition
GET {{search-url}}/knowledgesources/{{knowledge-source-name}}?api-version={{api-version}}
Authorization: Bearer {{token}}

```

--------------------------------

### Get knowledge base definition response (JSON) - Python

Source: https://learn.microsoft.com/en-us/azure/search/agentic-knowledge-source-how-to-fabric-data-agent

Example JSON response from getting a knowledge base definition, showing properties like name, knowledgeSources, and retrievalReasoningEffort.

```json
 {
   "name": "my-kb",
   "description": null,
   "retrievalInstructions": null,
   "answerInstructions": null,
   "outputMode": null,
   "knowledgeSources": [
     {
       "name": "my-blob-ks"
     }
   ],
   "models": [],
   "encryptionKey": null,
   "retrievalReasoningEffort": {
     "kind": "low"
   }
 }

```

--------------------------------

### Navigate to the quickstart folder

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-powershell

Changes the current directory to the quickstart folder for the keyword search sample.

```bash
cd azure-search-dotnet-samples/quickstart-keyword-search/AzureSearchQuickstart

```

--------------------------------

### Navigate to quickstart folder and open in VS Code

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-powershell

Changes to the quickstart directory and opens it in Visual Studio Code.

```bash
cd azure-search-rest-samples/Quickstart-keyword-search
code .

```

--------------------------------

### Get knowledge base definition response (JSON)

Source: https://learn.microsoft.com/en-us/azure/search/agentic-knowledge-source-how-to-fabric-data-agent

Example JSON response from getting a knowledge base definition, showing properties like Name, KnowledgeSources, Models, and ETag.

```json
 {
   "Name": "earth-knowledge-base",
   "KnowledgeSources": [
     {
       "Name": "earth-knowledge-source"
     }
   ],
   "Models": [
     {}
   ],
   "RetrievalReasoningEffort": {},
   "OutputMode": {},
   "ETag": "\u00220x8DE278629D782B3\u0022",
   "EncryptionKey": null,
   "Description": null,
   "RetrievalInstructions": null,
   "AnswerInstructions": null
 }

```

--------------------------------

### Example OneLake data source with system-assigned managed identity

Source: https://learn.microsoft.com/en-us/azure/search/search-how-to-index-onelake-files

Example data source definition using a system-assigned managed identity. Shows actual GUIDs and a folder name. The identity block is omitted.

```json
{    
  "name": "mydatasource",  
  "description": "description",  
  "type": "onelake",  
  "credentials": {  
    "connectionString": "ResourceId=a0a0a0a0-bbbb-cccc-dddd-e1e1e1e1e1e1"  
  },  
  "container": {  
    "name": "11111111-1111-1111-1111-111111111111",  
    "query": "folder_name"  
  }
} 

```

--------------------------------

### Run AgenticRetrievalQuickstart with dotnet run

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-agentic-retrieval

Executes the quickstart application to create an index, upload documents, configure a knowledge source and knowledge base, and run agentic retrieval queries. Run from the project directory containing the .csproj file.

```bash
dotnet run --project AgenticRetrievalQuickstart.csproj

```

--------------------------------

### Expected output from the quickstart application

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-powershell

The output shows the sequence of operations: deleting and creating the index, uploading documents, waiting for indexing, and running six queries. It includes sample results for each query, such as returning all documents, filtering and sorting, field selection, faceting, document lookup, and autocomplete.

```text
Deleting index...

Creating index...

Uploading documents...

Waiting for indexing...

Starting queries...

Query #1: Search on empty term '*' to return all documents, showing a subset of fields...

HotelId: 3
Name: Gastronomic Landscape Hotel
Rating: 4.8

HotelId: 2
Name: Old Century Hotel
Rating: 3.6

HotelId: 4
Name: Sublime Palace Hotel
Rating: 4.6

HotelId: 1
Name: Stay-Kay City Hotel
Rating: 3.6


Query #2: Search on 'hotels', filter on 'Rating gt 4', sort by Rating in descending order...

HotelId: 3
Name: Gastronomic Landscape Hotel
Rating: 4.8

HotelId: 4
Name: Sublime Palace Hotel
Rating: 4.6


Query #3: Limit search to specific fields (pool in Tags field)...

HotelId: 2
Name: Old Century Hotel
Tags: [ pool, free wifi, concierge ]


Query #4: Facet on 'Category'...

HotelId: 3
Name: Gastronomic Landscape Hotel
Category: Suite

HotelId: 2
Name: Old Century Hotel
Category: Boutique

HotelId: 4
Name: Sublime Palace Hotel
Category: Boutique

HotelId: 1
Name: Stay-Kay City Hotel
Category: Boutique


Query #5: Look up a specific document...

3
Query #6: Call Autocomplete on HotelName...

san
sarasota

Complete. Press any key to end this program...

```

--------------------------------

### Check PowerShell version

Source: https://learn.microsoft.com/en-us/azure/search/search-manage-powershell

Checks the installed PowerShell version. Run this before proceeding with the article's examples.

```powershell
$PSVersionTable.PSVersion

```

--------------------------------

### Get Index Statistics Response (JSON)

Source: https://learn.microsoft.com/en-us/azure/search/search-how-to-delete-documents

Example response from the Indexes - Get Statistics REST API, showing document count, storage size, and vector index size. Use this to confirm deletion later.

```json
{
  "documentCount": 12,
  "storageSize": 123456,
  "vectorIndexSize": 123456
}
```

--------------------------------

### Navigate to the quickstart folder

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-vector

Change directory to the vector search quickstart folder.

```bash
cd azure-search-dotnet-samples/quickstart-vector-search

```

--------------------------------

### Example response for shared private link status

Source: https://learn.microsoft.com/en-us/azure/search/search-indexer-howto-access-private

Example JSON response from the Shared Private Link Resources - Get API. The 'status' field shows the connection state, and 'provisioningState' indicates the resource provisioning status.

```json
{
      "name": "blob-pe",
      "properties": {
        "privateLinkResourceId": "/subscriptions/aaaa0a0a-bb1b-cc2c-dd3d-eeeeee4e4e4e/resourceGroups/contoso/providers/Microsoft.Storage/storageAccounts/contoso-storage",
        "groupId": "blob",
        "requestMessage": "please approve",
        "status": "Approved",
        "resourceRegion": null,
        "provisioningState": "Succeeded"
      }
}

```

--------------------------------

### Navigate to the quickstart folder

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-powershell

Changes the current directory to the quickstart-keyword-search folder within the cloned repository.

```bash
cd azure-search-java-samples/quickstart-keyword-search

```

--------------------------------

### Navigate to the quickstart folder

Source: https://learn.microsoft.com/en-us/azure/search/search-get-started-powershell

Changes the current directory to the quickstart-keyword-search folder within the cloned repository.

```bash
cd azure-search-javascript-samples/quickstart-keyword-search

```

--------------------------------

### Example OneLake data source with user-assigned managed identity

Source: https://learn.microsoft.com/en-us/azure/search/search-how-to-index-onelake-files

Example data source definition using a user-assigned managed identity. Shows actual GUIDs and a folder name. The userAssignedIdentity is the resource ID of the managed identity.

```json
{    
  "name": "mydatasource",  
  "description": "description",  
  "type": "onelake",  
  "credentials": {  
    "connectionString": "ResourceId=a0a0a0a0-bbbb-cccc-dddd-e1e1e1e1e1e1"  
  },  
  "container": {  
    "name": "11111111-1111-1111-1111-111111111111",  
    "query": "folder_name"  
  },  
  "identity": {  
    "@odata.type": "Microsoft.Azure.Search.DataUserAssignedIdentity",  
    "userAssignedIdentity": "/subscriptions/333333-3333-3333-3333-33333333/resourcegroups/myresourcegroup/providers/Microsoft.ManagedIdentity/userAssignedIdentities/demo-mi"  
  }  
}

```

--------------------------------

### GET response with redacted credentials

Source: https://learn.microsoft.com/en-us/azure/search/cognitive-search-custom-skill-web-api

Example of a GET response for a skill using header-based authentication and an Azure Function URI. The service returns <redacted> for httpHeaders values and ?code=<redacted> for query parameters to prevent credential exposure.

```json
{
  "@odata.type": "#Microsoft.Skills.Custom.WebApiSkill",
  "uri": "https://contoso.example.org/api?code=<redacted>",
  "httpMethod": "POST",
  "name": "myCustomSkill",
  "httpHeaders": {
    "Authorization": "<redacted>",
    "Ocp-Apim-Subscription-Key": "<redacted>"
  }
}
```
Tokens:
Raw