Clean AI™ within Clean & Match v11 can now automatically prepare your data for cleaning by intelligently identifying column types and applying the appropriate cleaning rules — saving time and reducing manual setup. 


How it Works

When a dataset is imported, Clean AI™ analyses:

  • Column names (e.g. “Address”, “Postcode”, “Email”, etc)
  • Data patterns within each column


Using this, it predicts the most likely column type and automatically configures the Cleaning Matrix with the appropriate rules.


Configuration Overview

This Clean AI™ Configuration screen allows you to control how columns are identified and how cleaning rules are applied.



1. Column Types

On the left-hand side, you’ll see a list of predefined column types such as:

  • Address
  • City
  • Email
  • Company
  • Phone

These represent the categories Clean AI™ assigns to your data.


2. Column Name Matching

For each column type, you can define how Clean AI™ recognises it.

You can add multiple possible column names (e.g. for Address: addr, address, street, mailing, shipping).

Each entry includes:

  • Lookup Type
    • Exact – must match the column name exactly
    • Contains – matches if the text appears anywhere in the column name
    • Fuzzy – allows approximate matches (useful for misspellings)
  • Precision
    • Used with fuzzy matching to control how close the match must be

This ensures flexibility when working with inconsistent or messy source data.


3. Cleaning Options

Each column type has predefined cleaning rules that can be automatically applied.

Examples include:

  • Removing leading/trailing spaces
  • Removing commas, dots, hyphens, apostrophes
  • Standardising formatting

When enabled:

  • If the Statistics panel detects an issue (e.g. trailing spaces present),
  • The corresponding rule is automatically selected in the Cleaning Matrix

If a rule is not enabled, it will not be applied automatically.


4. Pattern Recognition (Optional)

You can assign a Pattern Name to a column type.

This allows Clean AI™ to:

  • Identify valid and invalid values
  • Automatically populate the Pattern Statistics

Useful for structured fields such as:

  • Emails
  • Phone numbers
  • Postcodes


Creating Custom AI Rules

Clean AI™ is fully configurable to match your data environment.

To create a new column type:

  1. Enter a name under AI Type
  2. Click Add
  3. Define:
    • Column name variations
    • Lookup types (Exact / Contains / Fuzzy)
    • Cleaning rules to apply
  4. (Optional) Assign a pattern for validation

This allows you to tailor Clean AI™ to your organisation’s specific data formats.


Why This Matters

Clean AI™ removes the need to manually configure cleaning rules for every dataset.

Instead, it:

  • Automatically recognises your data
  • Applies the right cleaning logic
  • Speeds up project setup
  • Ensures consistency across datasets