Skip to content
Klarfolder

Sort a folder of documents, and see what is missing

Drop a whole folder. Klarfolder proposes a type for each document, a name, a place in the tree, and the list of what is missing against a mortgage, rental, onboarding or public tender file. Nothing is sent: it all happens in this tab.

Or drop a whole folder, a ZIP archive or a few files here. Nothing is sent: everything stays in this tab.

  • 14 tools, one workshop
  • 32 document types identified
  • 0 files uploaded

The tools

One address per tool, and the same workshop behind: each opens at the step it is about. Start with the one you were looking for.

Identify

Name

  • Rename a folder of filescommon

    A naming pattern applied to a whole folder, with the before and after preview and the undo.

  • Naming pattern

    The naming model, its tokens, and what it produces on your documents before writing.

Sort

  • Sort a folder into a treecommon

    By family, by person, by year, by issuer: the tree is proposed, never imposed.

  • Find duplicatescommon

    Byte-for-byte identicals, scans proposed as duplicates, and the series that are not.

Check what is missing

Produce

Why this one rather than another

Three differences, and they all come down to the same idea: never decide in your place.

  • Nothing leaves the tab

    Your documents are not uploaded: they are read by your browser. The page is not allowed to open a connection to another site, and you can cut the network to check.

  • The plan before the writing

    No file is renamed or moved before the whole table has been shown, row by row, with the reason for each name. And it all comes undone.

  • What is missing, not just what is there

    The question in front of a folder is not “what do I have”, it is “what is missing”. Five ready-made lists answer it, and each says where it comes from.

How it goes

Six steps, and you can stop at any of them.

  1. You drop a folder

    A whole folder, a ZIP archive, or a few files. Subfolders and archives are opened.

  2. The tool reads and proposes

    A type for each document, its date, its issuer, the person, and above all its confidence level.

  3. You correct

    What is wrong gets corrected, and the correction spreads to the group you choose. The number of documents affected is shown first.

  4. You see what is missing

    Against the list you choose, with the precise month when it is a series.

  5. You read the plan

    The name and folder of every document, before writing, with the reason for each.

  6. You apply, or not

    The copy is verified document by document. The list comes out as a PDF. And the sorting can be undone.

What Klarfolder does not do

A tool that promises too much ends up filing wrongly in silence. Here is what this one will not do, and it is not a provisional list.

  • It certifies no document genuine. It reads what is written; it checks no stamp, no signature, and no reference against any authority.
  • It does not say a file is acceptable. It says what is missing against a list you chose, and the list carries its source and its date.
  • It does not invent a name. A document it has not identified keeps its original name and goes to 00-a-classer/.
  • It deletes nothing. Even a certain duplicate is not erased: it is simply not copied, and the log says so.
  • It extracts no fields. Reading an amount, an account number, an address is Klarentry's job. Klarfolder reads in a document only what it takes to identify it.

Who it is for

Six professions spend their days on folders of documents. Each has its page.

  • Mortgage brokers

    The mortgage file, two borrowers, and the three statements that are always missing.

  • Lawyers

    The list of exhibits, the numbering, and a file made of screenshots and emails.

  • Notaries

    Deeds, civil status records, and a file arriving in bulk from three directions.

  • Human resources

    Ten arrivals a month, and four chasing emails each.

  • Accountants

    Clients' documents, named by the clients.

  • Letting agents

    Twenty applications, and what may not be asked for.

What has been measured

The figures on this page come from repeatable measurements, on a synthetic corpus and on real working folders.

MeasurementResult
Identification over 405 synthetic documents100 % correct type, zero document filed under the wrong type
Verified copy of 1,000 files5.5 seconds, 1,000 identical, undone in 0.7 seconds
A 300-page scanned stack, read35 seconds with six workers
Digests of 142 megabytes1.3 seconds

Start with your own folder

There is nothing to install, nothing to create, and no file will leave your computer.