Dexxter Migration
Import Dexxter yearly document exports into Workestra Finance expenses and invoices with vision-assisted review.
Use the Dexxter migration tool when you are moving accounting documents from Dexxter into Workestra Finance. The importer is built for Dexxter yearly document ZIP exports and turns reviewed documents into Finance expenses, Finance invoices, or archive-only records.
Every import is review-first. Workestra scans and extracts fields, but it does not create Finance records until you confirm the rows and click Create Finance records.
Where to find it
Open:
Settings -> Administration -> Import / Export -> Dexxter finance migrationDirect route:
/settings/administration/import-export/dexxterYou need a selected workspace before starting the import.
What you can upload
Upload one or more Dexxter yearly ZIP exports. The importer expects the document export structure Dexxter provides, especially these folders:
| Dexxter folder | Default Workestra target |
|---|---|
expenses/ | Finance expense |
invoices/ | Finance invoice |
| Other folders or unsupported files | Archive only / needs review |
Supported document types include PDFs and common image receipts. Empty ZIP files are handled as an import with no documents found instead of a broken upload.
Before you start
Prepare the import first:
- Export the yearly document ZIPs from Dexxter.
- Keep the ZIP files unchanged. Do not unzip and repackage them unless support asks you to.
- Choose the correct Workestra workspace.
- Confirm that the Finance module is enabled.
- If scanned PDFs or image receipts are common, configure the workspace AI/vision provider so Workestra can extract fields from images.
Import workflow
1. Upload exports
Click Choose Dexxter ZIP files and select one or more ZIP files. Each selected archive is shown before scanning.
Click Scan documents to start the preview. Workestra reads the archive, stores source documents privately, classifies each file, and creates a review queue.
2. Review the summary
After scanning, the summary shows:
| Metric | Meaning |
|---|---|
| Archives | Number of uploaded ZIP files |
| Documents | Total documents found inside the ZIPs |
| Expenses | Documents classified as expenses |
| Invoices | Documents classified as outgoing invoices |
| Needs fix | Rows that cannot be imported until required fields are corrected |
If Needs fix is greater than zero, open the affected rows and complete the missing fields before creating Finance records.
3. Review each document row
The review queue lists every document with its proposed target, source filename, counterparty, date, amount, currency, and confidence score.
Select a row to edit extracted fields such as:
- target entity
- vendor or customer name
- invoice number
- issue date, due date, or expense date
- total, subtotal, and tax
- currency
- description
Use Target entity to decide what happens to the source document:
| Target entity | Result |
|---|---|
| Expense | Creates a Finance expense with the source document attached |
| Invoice | Creates a Finance invoice from the confirmed fields |
| Archive only | Stores the source document in import history without creating a Finance record |
Click Save row after fixing a selected document.
4. Create Finance records
When all required rows are valid, click Create Finance records.
Workestra then creates the confirmed Finance records and reports how many rows were:
- created
- skipped
- errored
After the import, open Finance Expenses and Finance Invoices to review the created records.
How field extraction works
Workestra uses the best available extraction method per document:
| Document type | Extraction behavior |
|---|---|
| Text-based PDF | Reads embedded PDF text where available |
| Scanned PDF or image receipt | Uses the configured vision model when available |
| Unsupported or unclear document | Creates a review row that you can edit manually |
Vision extraction is helpful for screenshots, scanned receipts, and PDFs without embedded text. It still requires human review because supplier names, VAT labels, totals, and dates can vary between documents.
Duplicate handling
Dexxter imports use a skip-first duplicate strategy. Workestra checks source documents and Finance records before writing.
Duplicates can be detected from:
- source file checksum
- Dexxter folder and filename
- invoice number
- vendor, date, and amount combinations
When a duplicate is detected, the row is skipped or marked for review instead of overwriting existing Finance data silently.
Common fixes
| Problem | What to do |
|---|---|
| Upload button stays disabled | Select at least one .zip file. |
| Import says no workspace is selected | Switch to the correct workspace and retry. |
| Rows stay invalid | Open each row, fill required fields, and click Save row. |
| A document is classified incorrectly | Change Target entity to Expense, Invoice, or Archive only. |
| Image receipts have low confidence | Configure vision extraction or fill the fields manually. |
| A duplicate is skipped | Check whether the expense or invoice already exists in Finance. |
Data and audit trail
The importer keeps the original Dexxter document as import evidence and records the import action in the workspace audit trail. Imported source files are stored privately and scoped to the workspace.
Only confirmed rows create Finance records. Rows marked Archive only remain evidence for the migration without changing Finance totals.
Next steps
- Expenses - review imported supplier expenses
- Invoicing - review imported outgoing invoices
- AI Tools (Finance) - use the assistant after migration to inspect finance data