Getting started with business context in the Data Catalog

Note

Business context and semantic search is in preview for AWS Glue and is subject to change.

This tutorial walks you through creating a glossary, tagging assets, and using semantic search to discover data by meaning.

Prerequisites

• An AWS account with the AWS Glue Data Catalog configured in a supported Region.

• The AWS CLI installed and configured.

• At least one table registered in the Data Catalog.

• An IAM role or user with permissions for AWS Glue Data Catalog actions.

Attach the following IAM policy to grant the required permissions:

{
    "Version": "2012-10-17",
    "Statement": [{
        "Effect": "Allow",
        "Action": [
            "glue:Search", "glue:PutAsset", "glue:GetAsset", "glue:DeleteAsset",
            "glue:PutAssetType", "glue:GetAssetType", "glue:DeleteAssetType", "glue:ListAssetTypes",
            "glue:CreateGlossary", "glue:UpdateGlossary", "glue:GetGlossary", "glue:ListGlossaries", "glue:DeleteGlossary",
            "glue:CreateGlossaryTerm", "glue:UpdateGlossaryTerm", "glue:GetGlossaryTerm", "glue:ListGlossaryTerms", "glue:DeleteGlossaryTerm",
            "glue:AssociateGlossaryTerms", "glue:DisassociateGlossaryTerms",
            "glue:PutFormType", "glue:GetFormType", "glue:DeleteFormType", "glue:ListFormTypes",
            "glue:PutAttachment", "glue:DeleteAttachment",
            "glue:ListIterableForms", "glue:BatchGetIterableForms"
        ],
        "Resource": "*"
    }]
}

Step 1: Create a glossary and tag an asset

To create a glossary

Run the following command:

aws glue create-glossary \
  --name "Enterprise Data Glossary" \
  --description "Standardized business definitions for enterprise data assets."

Example output:

{
    "Id": "d7xm3np5rk2w9j",
    "Name": "Enterprise Data Glossary"
}

To create a glossary term

Replace the glossary identifier with the Id from the preceding output.

aws glue create-glossary-term \
  --glossary-identifier "d7xm3np5rk2w9j" \
  --name "Active User" \
  --short-description "A user with at least one login in the last 30 days." \
  --long-description "An account that has logged in at least once within the trailing 30-day window."

Example output:

{
    "Id": "c2fymbu18rtsx5",
    "GlossaryId": "d7xm3np5rk2w9j",
    "Name": "Active User"
}

To associate the term with an asset

Run the following command:

aws glue associate-glossary-terms \
  --identifier "arn:aws:glue:us-east-1:123456789012:table/mydb/sales_transactions" \
  --glossary-term-identifiers "c2fymbu18rtsx5"

Step 2: Search for data using business context

Use the Search API to find assets by business meaning.

aws glue search \
  --search-text "active users"

Example output:

{
    "Items": [
        {"Id": "c9vq7sh2fk4t2h", "AssetName": "Customer Sales Transactions", "AssetTypeId": "glue-table"}
    ]
}

To filter results by asset type:

aws glue search \
  --search-text "active users" \
  --filter-clause '{"AttributeFilter":{"Attribute":"assetTypeId","Operator":"equals","Value":{"StringValue":"glue-table"}}}' \
  --max-results 10

Using AI agents with the catalog

MCP-compatible AI agents can discover catalog assets, retrieve business context, and load skill content using skills from the AWS Agent Toolkit. You can get catalog skills in the following ways:

• Bundled with a plugin – Install the aws-data-analytics plugin, which includes a curated set of catalog skills available to the agent immediately after installation. For instructions, see Installing plugins in the AWS Agent Toolkit User Guide.

• Installed locally – Download individual skills from the Agent Toolkit for AWS repository on GitHub and add them to your agent's skills directory. The following skills support catalog workflows:

  • exploring-data-catalog

  • finding-data-lake-assets

Next steps

• Attach forms to standardize metadata fields such as data residency or retention policy.

• Create skill assets that provide AI agents with domain context for your data.