MazerikMazerik
Documentation home

Core Concepts

Transaction Normalization

Transform fragmented transaction text into stable, queryable records.

Last updated: 2026-03-24

Why normalization exists

Raw transaction text is inconsistent across banks, processors, countries, and channels. Normalization creates a reliable layer for analytics and decisioning.

Input and output model

  • Input: raw descriptor plus optional amount/date metadata.
  • Output: canonical transaction wording and structured signal fields.

Practical implementation pattern

  1. Keep original descriptor unchanged for audit.
  2. Store normalized descriptor as a separate field.
  3. Attach confidence and processing metadata.

Example

{
  "raw_description": "POS 09321 MKTPLACE*ONLINE",
  "normalized_description": "online marketplace purchase",
  "normalization_confidence": 0.94
}

Previous

Authentication

Open previous page

Next

Entity Intelligence

Open next page

Need help?

Talk with our team about architecture and rollout decisions.