Faasera Documentation

User Guide: Faasera Validation Service

The Faasera Validation Service ensures the accuracy, integrity, and structure of masked or synthetic data by comparing it against the original dataset or validation rules. It supports both rule-based and structural validation, helping teams verify that transformations have not compromised data utility or format compliance.

---

Why Use Validation?

Benefit	Description
Data Integrity Checks	Ensure key attributes retain relational integrity (e.g., FK/PK mappings)
Schema Conformance	Detects schema mismatches between source and masked data
Format Validation	Confirms the output matches required formats (e.g., SSN, DOB)
Quality Assurance	Flags anomalies or inconsistencies introduced during masking
Audit Traceability	Outputs validation logs for compliance and traceability

---

Modes of Validation

Faasera supports the following validation modes:

Mode	Description
STRUCTURAL	Validates column formats, nullability, regex patterns, and data types
SEMANTIC	Compares masked vs original value behavior or distribution
REFERENTIAL	Verifies foreign key and primary key relationships post-transformation
POLICY_BASED	Enforces explicit rules defined in your masking or data policy files

---

Configuration Example (Validation Policy)

Validation rules can be declared inline or in a separate validation policy:

{
  "validation": {
    "enabled": true,
    "rules": [
      {
        "type": "STRUCTURAL",
        "column": "email",
        "pattern": "^[\w.-]+@[\w.-]+\.\w{2,}$"
      },
      {
        "type": "REFERENTIAL",
        "keyType": "FK",
        "columns": ["user_id"],
        "referencedTable": "users",
        "referencedColumns": ["id"]
      }
    ]
  }
}

---

Example Output

After running validation, Faasera generates detailed reports:

{
  "table": "customers",
  "status": "PASSED",
  "validationSummary": {
    "totalRows": 10000,
    "failedRows": 23,
    "errorTypes": {
      "regexMismatch": 10,
      "nullViolation": 5,
      "fkMismatch": 8
    }
  }
}

These reports are stored in the Faasera UI and can be exported to PDF/CSV or accessed via API.

---

Best Practices

• Use STRUCTURAL validation for format-sensitive fields like IDs, emails, and dates.
• Use REFERENTIAL validation after masking datasets with foreign key dependencies.
• Integrate validation into CI/CD or data QA pipelines for continuous assurance.
• Combine with Audit Logs for full traceability and compliance readiness.

---

Integration Options

Environment	Integration Example
Airflow	Validation task after masking DAG step
Azure Data Factory	Add validation activity in a pipeline before publishing
Spark	Run validation SDK step after masking
Snowflake	Use UDFs or scripting to call Faasera SDK

---

Summary

The Validation Service ensures that post-masked or synthetic datasets retain structural, semantic, and relational integrity. It prevents compliance regressions and ensures that data quality is preserved after transformations — without requiring manual QA.

For more details, visit www.faasera.ai