When selecting the random text masking option, the output often results in a random string (e.g., 'adoowytkhusaosamnicgepexyspkbeuijmidxdxfukcbimaxk') that bears no resemblance to the format or structure of the original sensitive data. For fields like a Loan Name ('Application Id: 1234567 - Jones'), this loss of context (ID prefix, number, and name components) makes the masked data ineffective for systems testing that relies on that specific data structure.

We need an option for Custom or Pattern-Based Masking to allow users to define a template that maintains the field's structural components while ensuring the sensitive parts are de-identified. This would guarantee that masked data retains utility for testing purposes.