Using Synthetic Secure Lookup Lists

Overview

Synthetic Secure Lookup Lists generated with AI Services help organizations mask sensitive data with realistic, synthetic alternatives. This is especially valuable for test environments and compliance scenarios.

Supported data types and use cases

AI Services can generate synthetic values for a variety of commonly masked data types, including (but not limited to):

  • First names

  • Last names

  • Addresses

  • Medical conditions

These lists are intended for non-production use only and should not be used to simulate real individuals or generate live production data.

Generating a lookup list

To generate a synthetic list:

  1. Go to Compliance > Resource Files in DCT.

  2. Click Generate Synthetic List.

  3. Choose a pre-built template for the data type you want.

  4. Optionally, if the desired data type is not one of the pre-built templates, choose the Custom option and provide a custom prompt.

  5. Enter the desired number of records.

  6. Generate a preview to ensure it matches your expectations.

  7. Finalize and save the list.

    1. Full generation can take some time, potentially hours if thousands of values are requested.

If you select Allow Duplicates, the system will permit repeating values. If disabled, the system may return an error if it cannot generate enough unique values.

Reviewing and downloading the list

After generation, the list appears in the Resource Files section. Download the list in text format, which is compatible with the Delphix Continuous Compliance Engine.

Importing and using in the Compliance Engine

  1. Log into the Continuous Compliance Engine. Go to Settings > Algorithms.

  2. Choose to create a new algorithm and select Secure Lookup for Algorithm Type.

  3. Provide a name for the new algorithm.

  4. Upload the synthetic list as a custom lookup file.

  5. Adjust other settings as desired, then save the algorithm.

Once saved, the algorithm may be assigned to domains and masking rulesets as with any other algorithm.

Performance best practices

When using AI Services:

  • Avoid running generation tasks during system upgrades or peak periods.

  • Monitor system load and adjust usage as needed.

  • Running multiple synthetic list generation tasks simultaneously will increase the time needed to generate each list but may provide overall better throughput.

Data ownership reminder

All generated data belongs solely to you, the customer. Delphix does not store, retain, or access the generated content at any stage.

Troubleshooting data generation

If a generation task fails:

  • Confirm AI Services is enabled and the model is installed.

  • Review Admin > Operations for detailed error messages.

  • If the request requires unique values, reduce the number of rows or modify the content type and optional attributes selected.

For persistent issues, gather logs and contact the Delphix Support Portal.