Every file enters through the same gate and leaves as a searchable document.
Drag in files or whole folders, or let the scanner drop them into a watched folder. NG DMS scans every file for viruses, reads its format from the content and spots duplicates by fingerprint, then reads it: text per page, OCR in Serbian Cyrillic, Serbian Latin and English, and an archival version in PDF/A-2b.
- Files and whole folders
- OCR srp · srp_latn · eng
- PDF/A-2b archival version
- Duplicates by SHA-256
From a pile of files to orderly documents
Every file is a task with the outcome of each step, in words. Nothing is lost and nothing passes unchecked.
Files and whole folders
Drop anywhere in the application, subfolders included. Files above 20 MB go in 8 MB parts and resume after a dropped connection. A title template, type, tags and confidentiality can be set with the upload.
A gate before processing
Antivirus checks every file, and when the check is not possible the file is refused. The format is read from the content, not the extension, and compared with the formats the organization allows.
Duplicates by fingerprint
The SHA-256 is compared with the originals and archival versions of the whole organization, the trash included. The message never reveals the title of a document you cannot see; with the right, a file is entered anyway, linked as a duplicate.
Watched folders and two-sided scanning
The scanner leaves a file in a folder and the system takes it only once it stops changing. Subfolders become tags, odd and even pages merge into one document, and “Process once” brings in an existing folder.
OCR, text per page, language and date
Scans get text through OCR for Serbian in Cyrillic and Latin and for English. Text is kept per page, the language is detected and the document date is read from the content; when it is unclear, the field stays empty.
PDF/A-2b archival version
Scans, Office files and e-mails (EML, MSG) get an archival version in PDF/A-2b, with a first-page thumbnail. The original stays exactly as it arrived, and a signed PDF is never altered.
Barcodes, separators and ASB labels
A separator sheet splits a scanned stack into several documents, an ASB barcode sets the number of the paper original, and a tag barcode adds a tag. Separator sheets print as PDF, and ASB label sheets as PDF or ZPL.
Intake from a mailbox
Rules per e-mail account enter attachments or the whole message as a document, with the sender and date received. A single e-mail can also be entered as a file (EML, MSG).
SFTP intake and similar scans
Intake over SFTP/FTPS for scanners and systems that cannot write to a shared folder, recognising the same content even in a different scan.
What happens to a file
The same path for the browser, a watched folder, a scanner and e-mail. A document becomes visible only once it is saved, never half-made.
- 01
Intake
A file arrives through the browser or from a folder and becomes a task on the “Document processing” screen, with progress shown only to those allowed to see the document.
- 02
Checks
Antivirus, format by content and a SHA-256 fingerprint. An infected or disallowed file is refused, and a duplicate is reported before anything is created.
- 03
Processing
Text per page, OCR, language, date, barcodes, a PDF/A-2b archival version and a thumbnail. Each step records its outcome: done, skipped or failed, with the reason.
- 04
Document
Once saved, the document is in the list and in search. A failed task is retried, and a document can be reprocessed later without losing the original.
What changes in practice
Unchecked never passes
When the antivirus is unavailable, the file is refused with a message rather than let through “just in case”. The task is retried with one click once the check works again.
The original stays untouched
The archival version is added next to the original, not in its place. Downloads give the PDF/A by default, and the original is always there. If OCR is down, the document is still entered and its text follows later.
Empty beats wrong
An unclear language stays empty, a date before 1900 or in the future is not taken, and a damaged file is kept without text. A wrong value would cost more later than an empty field.
Frequently asked questions
Which formats does NG DMS accept?
PDF, images (JPEG, PNG, TIFF, GIF, BMP, WEBP, HEIC), text (TXT, CSV, Markdown), Office and OpenDocument files (Word, Excel, PowerPoint, RTF) and e-mails (EML, MSG). The format is determined from the content, and an organization administrator can narrow the list.
Does OCR read Serbian Cyrillic?
Yes. OCR works for Serbian in Cyrillic and in Latin script and for English. Text is kept per page, and search finds a Cyrillic document when you type in Latin, and the other way round.
What happens when the same file arrives twice?
The SHA-256 fingerprint is compared with the originals and archival versions of every document in the organization, the trash included. The “Document already exists” message names the document only if you can see it. A user with the right can enter the file anyway, linked as a duplicate.
Can a scanner send straight to NG DMS?
On an on-premises installation, yes, through a watched folder: the file is taken once the scanner finishes writing, subfolders become tags, and two-sided scanning merges odd and even pages. In the cloud the scanner sends to the organization’s mailbox or over SFTP.
Does processing change my original file?
No. The original is kept exactly as it arrived, and the PDF/A-2b archival version is created next to it. A signed PDF is never altered: its archival version is the original itself.
Related
- Classification and metadataTypes, tags, correspondents, attributes and counters that the organization maintains itself.
- Search and viewerCyrillic and Latin find each other, and a hit opens the exact page of the document.
- SaaS and on-premisesIn the cloud with data in the EU, or the same package on your premises — and a licence verified locally.
- Paper and scanning4 min read
ASB labels and separator sheets: how a paper original gets its number
A barcode label on the first page, a separator sheet between documents and a scanned stack that splits itself — so a scan always leads back to the binder.
- Processing4 min read
OCR for Cyrillic and Latin script and PDF/A: what processing reads from a scan
A scan in Cyrillic, a contract in Latin script, a letter in English — all three become searchable page-level text and a PDF/A-2b archival copy, while the original stays untouched.
Want to see how this looks in NG DMS?
We will show you in a demo, on examples from your own registry office — from the scan to NG Copilot’s answer. No obligation.
- office@konissoftware.com
- Phone
- +381 66 000 977