Skip to main content
Docs

Bank Import

Not sure about a term? See the Glossary.

What the Bank Import page is for

Bank Import is designed for initial onboarding with existing bank activity.

The page is split into clearly separated sections:

  • Upload for new CSV files
  • Import History for open and already executed uploads
  • Choose what to include for files, date range, and optional purpose
  • Imported transactions to review the currently visible rows
  • Apply data for the actual first import and its result summary

You can:

  • upload a bank CSV file
  • upload multiple CSV files for the same account together
  • store newly recognized transactions in an import list
  • apply a clearly selected date range and file selection
  • create historical payments and, for stable patterns, also plan items

Important:

  • Upload alone does not create payments.
  • Only the Apply data step starts the actual first import.
  • The process is intentionally clear and traceable and designed for onboarding.
  • The original file itself is not stored.
  • Valutra does not keep every transaction detail, only the information needed for import, duplicate detection, and assignment.
  • Account numbers are not processed.

How the process works

  1. Upload CSV
  2. Choose what to include: filter by file, date range, and optionally purpose
  3. Review imported transactions
  4. Apply data
  5. Review result and history

The important rules are:

  • only the currently visible and still unprocessed rows in your current selection are processed
  • only still-open uploads are available for a new import execution
  • import never starts automatically in the background
  • one uploaded import can only be executed once
  • existing data is not considered or merged; newly created payments and plan items are always added on top

The interface shows status, warning, and error information directly near the affected step. After upload and import execution, focus moves to the updated status or result area so the next step stays easy to continue by keyboard.

Upload and import list

Uploading stores:

  • import details about the uploaded file, not the file itself
  • only the information that is actually needed for imported transactions
  • newly recognized imported transactions in a reduced form

It does not yet create:

  • actual payments
  • changes to existing plan items

Important for privacy and safety:

  • the original CSV file is not stored permanently
  • not every detail from every row is kept
  • account numbers are not processed

The uploaded data remains visible in the import list so you can later trace which file was uploaded when and when it was actually applied.

If multiple CSV files belong to the same account, you can upload them first and process them together afterwards.

That is especially useful when recurring-series detection needs enough evidence across several files.

Once an uploaded file has produced business records, it becomes already imported:

  • it remains permanently visible in Import History
  • its summary becomes the final result for that file
  • it can no longer be deleted
  • it no longer participates in later processing scopes

What the first import creates

During the first import, the visible imported rows are evaluated.

Possible outcomes:

  • Standalone payment: when no stable recurring pattern exists
  • Plan item + payment(s): when regular payments with enough evidence are detected, for example rent or phone bills
  • Warning or explicit skip: when automatic assignment would not be reliable enough in the active Initial Import flow

New payments and plan items created during import initially use the neutral CSV-Import category. You can refine the categorization afterwards.

Regular payments and collisions

Valutra creates plan items for regular payments automatically only when the pattern is stable enough.

That means:

  • variable amounts can still be acceptable
  • the timing pattern must still remain clear
  • if multiple imported payments would occupy the same place in a regular payment pattern, Valutra does not assign them automatically

Such cases show up as warnings or are not automatically imported instead of silently creating incorrect data.

Overlap with existing data

If your current selection overlaps existing payments or plan items, Valutra requires explicit confirmation before execution.

This matters because the first import does not silently clean up or merge existing data.

If you continue anyway, new payments and plan items are created in addition to what is already there.

What Bank Import intentionally does not do

The current process is for the first import only.

It does not include:

  • no ongoing sync
  • no background matching process
  • no preview-only mode
  • no undo for already created financial data
  • no deletion of files that were already applied

Once business records were created, the import entry remains available for history and duplicate detection.

Important:

  • files that were already applied can no longer be deleted
  • files that were already applied also cannot be executed again
  • there is no undo for created payments or plan items
  • if you want to remove an import afterwards, this can only be done manually by deleting the individual payments and plan items

Warnings, collisions, and failed rows are summarized in the result area. The page keeps those explanations directly visible instead of relying on hidden duplicate helper text.

If there is hardly any other data besides the import yet, a full account restart can be easier than cleaning this up manually.

But that also removes other existing data, for example:

  • bank accounts
  • loans
  • assets
  • projects and contracts
  • other financial data already created

Reading the import history

For each uploaded file, the import history shows items such as:

  • upload time
  • latest actual import execution
  • number of created payments
  • standalone payments
  • created plan items
  • skipped rows, warnings, and collisions

Open uploads and already imported files are intentionally separated: only open uploads can still be added to the active processing scope.

This lets you distinguish between:

  • file was uploaded
  • file was actually turned into payments and plan items

And for the active workflow:

  • open uploads can still be processed
  • already imported files remain journal history only

Next topics: