Documni
Menu
English
Join the beta list

Document recognition & reconstruction

Turn a scan into an editable document.

Change the words, replace a logo, move a picture or update a table. Documni rebuilds the content and layout of a document so you can work with each part.

Rebuild the page. Then make your changes.

Keep the design you need and change the parts you don’t. You don’t have to recreate the whole document by hand.

Text you can rewrite

Correct a name in a letter or change the headline on a poster.

Words become editable text blocks, with their position and formatting.

Edit the text itself

Logos and images you can replace

Swap an old logo, move a photograph or resize an illustration.

Visual elements become separate objects you can select and arrange.

Work with individual images

A layout you can adapt

Reuse an event poster with a new date, photograph and arrangement.

Reconstruction brings text, images and shapes together in their places on the page.

Keep or change the design

Table cells you can change

Update a price list or correct a quantity in a report.

Rows, columns and values are rebuilt as a table. Check that each value sits under the right heading.

Edit individual cells

The Documni reconstruction engine

From a picture of a page to the document itself.

Recognition identifies the page’s contents. Reconstruction rebuilds their structure and layout, so a scan becomes a document you can edit.

  1. Recognise

    Find the text, logos, images, shapes and tables on the page.

  2. Rebuild

    Recover text blocks, separate visual elements, table cells and their placement.

  3. Review and edit

    Compare the result with the original, correct anything that needs attention and make your changes.

A clear source makes a better starting point.

Capture the whole page

Keep the page flat and evenly lit, with small text and image edges in focus.

Check the details

Review names, numbers, image crops and table alignment against the original.

Keep your original

Some details may need manual correction. Keep the source so you can refer back to it.

Capture a document on your phone

Bring your documents to the beta.

Beta launches 25 November 2026. Join the list for the release email and your invitation. Taking part in testing is optional.