Below is an example of a typical workflow YAML file. It includes steps for masking and subsetting, starting from a backup file, but could easily be modified to subset from a live database.

version: 1

vars:
  # Path to the original backup file
  backup-path: ${env.backup-path}
  # Name of the database to restore the original backup to
  restored-database-name: ${env.restored-database-name}
  # Name of the database to subset to (must be different from the restored-database-name)
  subset-database-name: ${env.subset-database-name}
  # Path to where you want the final backup to be saved
  output-backup-path: /tmp/${env.output-backup-file-name}
  # Should we overwrite the existing backup file in this location?
  overwrite: true

connections:
  # Your server's connection string (needs SA permissions)
  server:
    ado: ${env.connection-string}
    ado-engine: ${env.connection-engine} # SqlServer or similar

jobs:
  job1:
    steps:
      # Restore the backup to the server
      - uses: deploy-backup
        with:
          backup-path: ${var.backup-path}
          connection: ${connection.server}
          database-name: ${var.restored-database-name}
          ephemeral: true
          output: =connection.restored-backup
      # Create a script that will create an empty schema of the database
      - uses: create-schema-script
        with:
          connection: ${connection.server}
          database-name: ${var.restored-database-name}
          output: =var.schema-script-file
      # Deploy the schema to the server as a new database
      - uses: deploy-from-script
        with:
          connection: ${connection.server}
          database-name: ${var.subset-database-name}
          ephemeral: true
          script-file: ${var.schema-script-file}
          output: =connection.subset-database
      # Subset the data from the restored database to the new database
      - uses: subset
        with:
          source-connection: ${connection.restored-backup}
          target-connection: ${connection.subset-database}
          output-file-path: =var.subsetting-file
      # classify the data
      - uses: classify
        with:
          connection: ${connection.subset-database}
          output-file-path: =var.classification-file
      # map classification data for masking
      - uses: map
        with:
          connection:  ${connection.subset-database}
          classification-file-path: ${var.classification-file}
          output-file-path: =var.masking-file
      # Mask the data
      - uses: mask
        with:
          connection: ${connection.subset-database}
          input-masking-file-path: ${var.masking-file}
      # Export the masked and subset data to a new backup file
      - uses: export-backup
        with:
          backup-path: ${var.output-backup-path}
          connection: ${connection.server}
          database-name: ${var.subset-database-name}
          overwrite: ${var.overwrite}
          output-file-path: =var.backup-file