Classifications

Classifications in Paradime are powered by dbt™'s meta configuration, which allows you to add custom metadata to your models, columns, sources, and other dbt™ resources. All classifications must be defined within the meta config block in your dbt™ YAML files. These classifications appear in the Paradime catalog as searchable and filterable metadata, helping with data governance and asset discovery.

---

Legacy vs Modern Syntax

Important: dbt™ 1.10+ requires meta configurations to be nested within config blocks. The legacy syntax shown below is deprecated and will be removed in future versions.

Deprecated (will be removed in future dbt™ versions):

models:
  - name: my_model
    meta:                    # Direct meta placement - deprecated
      classification: "restricted"
    columns:
      - name: my_column
        meta:                # Direct meta placement - deprecated
          pii: true

Current (dbt™ 1.10+ compatible):

models:
  - name: my_model
    config:
      meta:                  # ✅ Meta nested in config block
        classification: "restricted"
    columns:
      - name: my_column
        config:
          meta:              # ✅ Meta nested in config block
            pii: true

---

How Classifications Appear in Paradime

Once you've added meta configurations to your dbt™ files, they appear in two key places in the Paradime catalog:

1. Search and Filter Bar

At the top of the catalog home screen, click the Classification dropdown to search and filter by any custom metadata you've added. This allows you to quickly find assets with specific classification values like sensitivity levels, data domains, or ownership information.

2. Asset Details Panel

When viewing any data asset, classifications appear in the right-side panel under the Classification section. This displays all the meta key-value pairs you've defined, such as:

• is_sensitive: true

• owner: John

• Any other custom metadata you've configured

---

Where to Set Classifications

Classifications are always defined within the meta configuration block at two levels in your dbt™ project:

Model Level Classifications

Add classifications to entire models using the meta config block in your properties file:

models:
  - name: customer_orders
    description: "Customer order data with purchase history"
    config:
      meta:                  # Meta nested in config block
        classification: "restricted"
        data_domain: "sales"
        owner: "analytics-team"
        pii_contains: true
    columns:
      - name: customer_id
        description: "Unique customer identifier"
      - name: order_total
        description: "Total order amount"

Column Level Classifications

Add classifications to specific columns using the meta config block within column definitions:

models:
  - name: customer_orders
    description: "Customer order data with purchase history"
    columns:
      - name: customer_id
        description: "Unique customer identifier"
        config:
          meta:                  # Meta nested in config block
            classification: "public"
            pii: false
      - name: customer_email
        description: "Customer email address"
        config:
          meta:                  # Meta nested in config block
            classification: "restricted"
            pii: true
            retention_period: "7_years"
      - name: order_total
        description: "Total order amount"
        config:
          meta:                  # Meta nested in config block
            classification: "internal"
            currency: "USD"

---

Common Classification Examples

Here are some typical classification patterns teams use within the meta config block:

Data Sensitivity Levels

models:
  - name: my_model
    config:
      meta:                  # Meta nested in config block
        classification: "public"      # Safe for wide access
        # OR
        classification: "internal"    # Company employees only
        # OR
        classification: "restricted"  # Limited access required
        # OR
        classification: "confidential" # Highly sensitive data

Data Domains

models:
  - name: my_model
    config:
      meta:                  # Meta nested in config block
        data_domain: "finance"
        # OR
        data_domain: "marketing"
        # OR
        data_domain: "product"
        # OR
        data_domain: "operations"

Data Governance

models:
  - name: my_model
    config:
      meta:                  # Meta nested in config block
        owner: "data-engineering"
        pii_contains: true
        retention_period: "5_years"
        compliance_requirement: "gdpr"

---

Best Practices

1. Be Consistent: Use standardized classification values across your project

2. Document Standards: Define what each classification level means for your organization

3. Use Meaningful Keys: Choose descriptive meta keys that make sense to your team

4. Combine Classifications: Use multiple meta fields to provide rich context