Introduction

n today’s data-driven world, finding the right information quickly can make or break productivity. That’s where Azure AI Search comes in. As a cloud-based search-as-a-service solution, it enables organizations to index, enrich, and search across structured and unstructured content using AI.

This article is the first in a practical step-by-step series on Azure AI Search setup. Here, we’ll walk through the basics—creating a search service, preparing your data, indexing it, and performing your first search query.

By the end, you’ll have a working search engine ready to be integrated into your applications.

Step 1: Creating the Azure AI Search Service (Prerequisites)

Before setting up the service, make sure you have:

• An active Azure subscription.
• A resource group where the service will live.
• A data source (e.g., Azure Blob Storage, SQL, or Cosmos DB) you want to search.

Creating the Search Service

1. Log in to the Azure Portal.
2. Click Create a resource → AI + Machine Learning → Azure AI Search.
3. Choose a unique name for your search service.

1. Select a region close to your data for lower latency.
2. Pick a pricing tier (Free for learning, Standard for production).
3. Configure replicas (for query load) and partitions (for data size).
4. Click Review + Create to provision the service.

✅ You now have a cloud-hosted search service ready to be configured.

Step 2: Creating Blob Storage and Uploading Documents

Since we need a data source, let’s create Azure Blob Storage to hold your documents.

Creating a Blob Storage Account

1. In the Azure Portal, click Create a resource → Storage → Storage account.
2. Enter a unique name (e.g., mystorageaccountdocs).
3. Select the same region as your AI Search service for performance.
4. Choose Standard performance and Hot access tier.
5. Click Review + Create.

Creating a Blob Container

1. Open your Storage account → Containers → + Container.
2. Name it something like documents.
3. Set Public access level to Private (secure by default).

Uploading Documents

1. Open the documents container.
2. Click Upload and select a few test files:
   • PDFs (reports, contracts, manuals)
   • DOCX/TXT files (articles, notes)
   • Images (scanned documents for OCR)
3. Verify that files appear in your container.

✅ Your Blob Storage is now populated with documents, ready to be indexed.

Step 3: Preparing the Data Source

Next, we’ll tell Azure AI Search how to read from Blob Storage.

1. Go to your Search service → Data sources → Add.
2. Choose Azure Blob Storage as the source type.
3. Provide a name (e.g., docs-source).
4. Select your Storage account and documents container.
5. Use Managed Identity authentication so the Search service can securely access storage.

✅ Now the service knows where to pull content from.

Step 4: Creating the Indexer

The indexer connects your data source with an index (searchable schema).

1. In your Search service, go to Indexes → Add.
2. Define fields like:
   • id (string, key)
   • content (searchable text)
   • metadata_author (filterable, facetable)
   • metadata_storage_name (retrievable file name)
3. Save the index.

Then, go to Indexers → Add:

1. Select the data source you created earlier.
2. Choose the index you just built.
3. Name it docs-indexer.
4. Run the indexer manually the first time.

✅ The indexer will read your documents, extract text, and load searchable content into your index.

Step 5: Searching the Indexed Content

Using Search Explorer

1. Go to your Search service → Indexes → Search Explorer.
2. Enter a search query like:
   • report
   • “financial summary”
   • metadata_author:John Doe

Conclusion

In this first part of the Azure AI Search setup series, you:

• Created a Search service.
• Built a Blob Storage account and uploaded documents.
• Prepared a data source.
• Created an indexer to index your documents.
• Searched the indexed content with Search Explorer and APIs.

This setup forms the foundation of intelligent search applications. In the future article, we’ll cover designing indexes, using AI enrichment (OCR, key phrase extraction), and enabling semantic search.

Happy Searching…