Document digitisation and OCR
The paper archive sitting in cabinets becomes data you can search in seconds.
Find what was previously unfindable
The paper archive becomes full-text searchable by keyword, document number and issuing authority.
Preserve records before they degrade
Content is stored digitally and survives the deterioration of the paper original.
Feeds the other products
Digitised documents become source data for retrieval, report consolidation and reconciliation.
Paper archives are an asset and a burden
Many organisations still hold thousands of boxes of paper: dispatches, decisions and forms accumulated over years. They contain important information that is effectively unusable.
- Finding an old file means searching by hand, sometimes for hours
- Paper degrades, ink fades, and material can be lost permanently
- Scanning to image files still leaves the content unsearchable
- Vietnamese diacritics defeat many recognition tools
What changes after deployment
Documents are scanned or photographed and uploaded. The software rotates, crops and cleans the image before recognition, reads Vietnamese text with diacritics even on mid-quality scans, then classifies by document type, subject area, unit and date. The result is a PDF that keeps the original image but gains a text layer, so full-text search works.
What you gain
Convert paper documents into digital data, then automatically recognise, classify and store them centrally so they can be found when needed.
Input data
- Records, documents, dispatches, decisions and forms
- Scans from a document scanner
- Photographs of documents
- Image-only PDFs with no text layer
- Older archived material of uneven print quality
Core functions
- Recognise Vietnamese text with diacritics, including mid-quality scans
- Recognise tables and figures within documents
- Automatically rotate, crop and clean images before recognition
- Automatically classify by document type, subject area, unit and date
- Add a searchable text layer to PDF files
- Manage document versions and history
- Full-text search by keyword, document number or issuing authority
Processing flow
- 1Paper document
- 2Scan or upload images and PDFs
- 3Optical character recognition
- 4Classification and metadata tagging
- 5Central storage
- 6Search and retrieval
Outputs
- PDF files with a searchable text layer
- Digital text ready to feed the knowledge repository
- Metadata: document type, reference code, issuing authority, date
- A digital archive that can be searched in full text
- A list of pages to re-check where image quality was poor
Governance rule
Recognition accuracy depends on scan quality. The system flags low-confidence pages for an officer to re-check rather than quietly accepting a wrong result.
Frequently asked questions
- What accuracy percentage do you achieve?
- We do not quote a single number, because it depends heavily on scan quality, typeface and the condition of the paper. Our approach is to run a trial on your own documents and report the figure measured on that set.
- Can it read handwriting?
- Handwriting is considerably harder than print and results are less consistent. For records containing handwritten sections we recommend evaluating a real sample before committing to scope.
- Do we need to buy new scanners?
- Not necessarily. The software accepts files from your existing scanners and photographs. For large volumes, a high-speed scanner saves considerable time.
Part of AI assistant for government agencies
See it run on your own documents
Every solution sounds good in a description. The only way to know whether this one works for you is to run it against your real documents, templates and workflows. That is exactly the kind of demo we do.
The demo is free and carries no obligation. If it turns out we are not the right fit, we will say so.