Skip to main content
Question

AI Extraction Table Overview

  • July 7, 2026
  • 1 reply
  • 61 views
N

Hi everyone, I’m looking for the best approach to using AI Extraction to generate a well-structured table overview when extracting data from a large volume of documents.

At the moment, every time I extract the data, the extracted results are listed one below the other in Excel, document by document, which makes the output quite difficult to review and analyze.

Do you have any tips or recommended workflows to make the extracted data from AI Extraction more organized ?

    1 reply

    P

    Hi ​@Nader Jebali, 

    Below are some tips from our team. Let me know if this helps!
     

    Don't mix single-value fields and table/list fields in the same extraction.
    The stacked output happens when one extraction contains both header fields (one value per document) and line items (many rows per document). DataSnipper can't fit both on one row, so it stacks everything.
    So split them:

    • Header/summary fields (account number, dates, totals): keep as standard single fields, not under a List/Table key. You get one clean row per document, one column per field.
    • Line items (transactions, expense lines, status logs): extract as a Table in a separate extraction, each with a Document column to trace back.

    Add a short description to each field to keep columns consistent across layouts.

    Terms & ConditionsAccessibility statement