Auto-Categorization and Auto-Matching
  1. Help Center
  2. Reconciliation Center
  3. Auto-Categorization and Auto-Matching

Wait For Approval Before Categorization

This feature introduces a structured interface, ensuring a seamless and intuitive categorization process. Users will easily differentiate between transactions needing attention and those already reviewed, optimizing workflow efficiency.

Overview

The "Wait for Approval Before Transaction Categorization" feature enhances user oversight and control over Docyt AI's transaction categorization process. By introducing an approval step, this feature ensures greater accuracy, minimises errors, and builds user trust.

With this update, AI will no longer automatically categorize transactions without user input. Instead, transactions will go through a structured review and approval process before being pushed to the ledger.

How It Works

  1. AI Suggestion – The system pre-populates a category based on AI memory and past transactions.
  2. User Review – Users can verify, modify, or approve the suggested categorization before finalizing it.
Approval & Ledger Entry – Once approved, the transaction is categorized and pushed to the ledger.

 

Limitations of AI Transaction Categorization

AI categorization can sometimes lead to inaccuracies, particularly when it relies on past transactions that may contain errors. If previous entries are incorrect, there is a risk that future transactions will also be misclassified. This can necessitate manual re-categorization by users, which can be both inefficient and frustrating.


AI categorization based on past transactions can sometimes lead to incorrect entries being pushed to the ledger. If previous errors exist, future transactions may be categorized inaccurately, forcing users to manually re-categorize them—resulting in inefficiency and frustration.

Primary Advantage

This feature empowers users by allowing them to review AI-predicted categorizations before they affect financial records. Instead of automatic categorization, users have full control to validate or adjust transactions, ensuring accuracy before they are finalized in the ledger.

User Experience & Workflow

Transactions will be segmented into two categories:

  1. Pending Review: Transactions that have been suggested by Docyt AI but still need user validation. These will feature a "Categorize" call-to-action (CTA), encouraging users to assess and confirm the suggested categorizations.

  2. Awaiting Categorization: These are transactions that do not have AI-generated categorizations and require user input. They will feature a "Review" call-to-action (CTA), prompting users to manually assign categories and approve them before they enter into the ledger.