Sometimes AI extraction gets almost everything right - but you may still want to quickly fix a date, amount, field value, or transaction entry before exporting your data or sending it to an integration.
With Parseur’s Manual Data Editing feature, you can update parsed results directly inside the app without needing to reprocess the document or edit exported CSV/XLS files afterward.
This article explains when to use manual edits, what you can change, and how the editing tools work.
When Should You Use Manual Editing?
Manual editing is useful when:
A single field was extracted incorrectly
OCR misread a value
You want to adjust transaction data before exporting
A document has unusual formatting
You need a quick correction without changing your parser template
You want to fix one specific document only
Manual edits only affect the current document.
If you want to improve extraction results for future documents as well, you should update your fields/template configuration instead.
Opening the Parsed Data Editor
Open any processed document in Parseur:
You’ll see two tabs on the right side:
Data
Fields
The Data tab displays your extracted JSON structure and allows direct editing.
The Fields tab shows your parser field configuration and table structure.
Editing Parsed Data
To manually update parsed data:
Open the processed document
Go to the Data tab
Click the Edit data option from the Options & Download menu:
You can then directly modify values inside the parsed JSON structure:
The editor supports:
Editing field values
Modifying objects
Editing arrays/lists
Inserting new elements
Removing elements
Restructuring data types
Searching through the output for a specific value
After making changes, click Save.
You can also choose whether to:
Save the edit only
Or trigger integrations/webhooks again using the Trigger export to integrations option
Exporting Edited Data
Once edited, your updated data can be downloaded in multiple formats.
Available export formats include:
Excel (.xlsx)
CSV
JSON
You can also copy the parsed JSON directly from the editor.
Understanding the Data Structure
Parseur organizes extracted information using JSON structures.
Depending on your parser setup, your data may contain:
Structure | Description |
Value | A single piece of data |
Object | A group of related fields |
Array/List | A repeating collection of items |
Table | Structured rows and columns |
Example:
{
"account_name": "MR JOHN DOE",
"transactions": [
{
"date": "03 Jul",
"withdrawal": 6
}
]
}
Available Editing Actions
Edit a Value
You can directly modify text, dates, numbers, or extracted values.
Examples:
Correct a customer name
Fix an invoice total
Adjust a transaction amount
Remove Data
You can remove:
Individual fields
Array items
Objects
Entire sections
Useful when unwanted information was extracted:
Insert New Data
You can insert:
New fields
New objects
New array items
This is useful when information was missed during extraction:
Convert Data Types
The editor allows converting between structures such as:
Value → Object
Object → Array
Array → Object
This helps reorganize parsed data when needed.
Editing Tables and Lists
For repeating data such as:
Transactions
Line items
Products
Expenses
Parseur stores them as arrays/lists.
You can:
Add rows
Remove rows
Edit values inside rows
Reorganize entries
Triggering Integrations After Editing
If your document is connected to:
Zapier
Make
Webhooks
Google Sheets
CRM systems
APIs
You can optionally resend the updated data after editing - simply keep Trigger export to integrations enabled before saving:
Important Notes
Manual edits do not retrain the AI
Changing parsed data does not improve extraction for future documents.
To improve future parsing:
Update your fields
Adjust AI prompts
Modify your parser setup
Edits are document-specific
Changes apply only to the document you edited - other processed documents remain unchanged.
Using the Search Function
The parsed data editor also includes a built-in Search feature to help you quickly locate specific fields or values inside large JSON datasets.
This is especially useful when working with:
Long bank statements
Large invoice batches
Transaction-heavy documents
Complex nested data structures
To use it:
Open the Data tab
Click inside the search bar at the top of the editor
Type a field name or value:
Parseur will highlight matching entries within the parsed data structure, making it easier to:
Find specific fields quickly
Jump to values that need correction
Navigate large arrays and tables
Verify extracted information faster
Examples:
Search for
statement_end_dateSearch for a customer name
Search for a transaction amount
Search for a specific field path
This can significantly speed up manual review and correction workflows for larger documents.
Frequently Asked Questions
Can I edit parsed data after export?
Yes. You can reopen the document, edit the data, and export it again.
Will editing parsed data retrain the parser?
No. Manual edits only affect the current document.
To improve future extractions, update your parser configuration instead.
Can I undo changes?
If changes are saved, the edited version becomes the current document data.
(You may want to duplicate or export data first before making major edits.)
Can I edit table rows and transaction lists?
Yes. Arrays and table-like structures can be edited, expanded, or reorganized.
Can integrations receive the updated data?
Yes. Enable Trigger export to integrations before saving to resend the updated data.
If I reprocess the document I just edited, will the changes stay?
All parsed data including manual edits will be thrown out when a document is queued for reprocessing.