Successful Projects in Document Automation

Dutch municipalities face the challenge of quickly and securely processing bank statements from various banks (ABN Amro, ING, Rabobank, and SNS). We have successfully solved the core of this challenge with our Financial Search technology, which automatically recognizes documents, extracts relevant data, and categorizes transactions using a dynamic system.

The unique aspect of this approach is that categories contain no personal data and—if desired—can be shared with other participating municipalities, creating a network of best practices that are widely applicable.

The implementation also included a service desk portal, user logging, and extensive support. Thanks to this integrated approach, participating municipalities can organize their financial processes more efficiently and securely, maintaining full privacy and compliance.

Learn more about our cloud services

A successful project involved automating the invoice processing workflow for a mid-sized Dutch company. The goal was to minimize manual processing of incoming PDF invoices and increase administrative efficiency.

Incoming PDF invoices are automatically recognized at line level. Invoices that contain an XML file are immediately forwarded to the desired server location. Recognized invoices are placed via a secure connection in a folder directly accessible to the client’s ERP system.

Read more about invoice recognition

For a leading organization in the healthcare sector, we executed an innovative project focused on improving the recognition of hard-to-read documents. A significant portion of submitted documents could not be automatically recognized by existing OCR systems.

We developed an advanced image enhancement module based on machine learning and OpenCV. This module corrects perspective, restores distortions, and increases the resolution of documents, preparing them optimally for automatic recognition.

Discover our OCR technology

For an internationally operating organization, we carried out a project focused on automating complex invoice processing. The client faced challenges such as high dependence on manual processing, susceptibility to errors, and limited flexibility.

We implemented an advanced solution that automatically recognizes invoices and extracts relevant data—such as suppliers, amounts, VAT rates, and currencies—directly from documents. This was achieved through a combination of template recognition and machine learning.

For a government organization, we executed a project entirely focused on modernizing the processing of incoming mail and documents. The challenge was to efficiently process both paper and digital mail streams, ensure privacy, and realize integration with existing case management systems.

We implemented a solution that allows both physical and digital documents to be centrally scanned, recognized, and automatically provided with relevant metadata. A distinction is made between technical and content metadata, tailored to organizational requirements.

Learn more about process automation

For a major player in the telecom and technology sector, we executed a project focused on optimizing the scanning and recognition process for various documents, such as identity cards, bank cards, receipts, and business registry extracts.

Using our EasyData OCR Technology, documents such as passports, ID cards, driver’s licenses, and bank cards are automatically recognized and relevant data extracted. For European identity documents, the MRZ (Machine Readable Zone) is used as the basis for reliable and fast data extraction.

For a leading company in the manufacturing industry, we realized an advanced project aimed at automating technical document processing and performing complex calculations based on drawings. The challenge was to automatically extract data from weld drawings, validate it, and directly translate it into accurate calculations for production and quality control.

Using machine learning, specific weld symbols and relevant parameters (such as material type, plate thickness, weld length, and angle) are automatically recognized and extracted from the drawings. Missing or unclear fields are flagged, allowing the user to easily supplement or correct them via a user-friendly web portal.

Discover technical document recognition

For an international industrial organization, we executed a project focused on automatically reading and structuring complex certificate documents, such as material certificates for steel and other metals. The challenge was that these documents were semi-structured: tables and data appeared in various forms and layouts.

To improve extraction quality, reference lists were used for material types, "Heat No." formats, and diameter ranges. This minimizes errors in reading and validating values and allows deviations to be detected immediately.

For an archive institution dealing with large quantities of historical documents—including cards with handwritten notes, files, and inventory lists—we executed a project focused on automating the recognition and digitization of both handwritten and typed texts. The challenge was the enormous variety of handwriting, pencil notes, and varying scan and paper quality.

For documents with faint pencil notes or low scan resolution, image enhancement modules were applied. The recognized text was used directly to index documents and add metadata, making them searchable in the archive system.

Learn more about handwriting recognition

For an international financial services provider, we executed a project focused on fully automating the processing of various documents, such as contracts, invoices, and client forms. The organization dealt with large volumes of documents arriving daily through different channels (email, uploads, cloud folders).

We set up a workflow where documents are automatically retrieved from various sources, recognized, categorized, and relevant data extracted using AI-driven OCR and machine learning technology. The extracted data is automatically validated and sent directly to backend systems.

Discover full process automation

For a government organization, we set up an innovative project focused on automating the processing and validation of certificates and trade documents. The challenge was to quickly and reliably distinguish between genuine and fraudulent documents, recognizing and analyzing both handwriting and machine-written text.

Using our proprietary OCR technology and handwriting recognition modules, documents are read accurately, with missing or unclear data automatically flagged. Documents are classified and validated using reference lists and metadata.

Learn more about secure cloud solutions

For a trade and distribution company with a high volume of order processing, we executed a project focused on automating and optimizing the order process. The existing solution was no longer adequate: important order details were lost and much manual checking was required, leading to delays and errors.

The system is trained to recognize essential data such as delivery address, customer number, email, IBAN, phone numbers, order numbers (both dealer and end-customer), and tables with quantities and item numbers. Unknown or incomplete data is automatically supplemented via a database lookup.

Discover cloud-based order processing

For a technology company specializing in data protection, we executed a project around the development and implementation of an advanced solution for document anonymization. The client wanted to respond to the growing demand for privacy protection when processing large volumes of sensitive documents.

We fully integrated the existing anonymization platform with a scalable cloud environment. As a result, large numbers of documents can be processed quickly and securely, with page-by-page processing running in parallel for maximum speed and capacity.

Learn more about privacy protection

For a national organization responsible for providing insight into large volumes of annual accounts and financial reports, we executed a project focused on converting and structuring unstructured financial documents. The challenge was to quickly and accurately extract relevant financial data from thousands of annual reports.

Using smart algorithms, consolidated balance sheets and other key data are automatically recognized, checked for completeness, and compiled into a clear Excel file or database. The export can be fully tailored to the client’s wishes and is suitable for further analysis or monitoring.

Discover financial data extraction

For a large Dutch theme park, we executed a project focused on optimizing the findability and accessibility of a large technical archive. The challenge was to make thousands of technical drawings, maintenance documents, and inventory lists more searchable for technical staff.

We started with a thorough inventory of the archive material from multiple attractions. This included reviewing the structure of existing folders, the presence of inventory lists, and the quality of technical drawings and maintenance documents. For documents with poor scan quality, image enhancement was applied.

Discover archive digitization

For a logistics service provider active in international trade, we realized a project focused on automating the processing and checking of transport and freight-related documents, such as CMRs, loading and unloading forms, and accompanying certificates. The client wanted to drastically reduce manual checks and minimize compliance risks.

All submitted transport documents are automatically recognized and relevant data—such as PO/SO number, truck and trailer information, loading and unloading locations, and product details—are extracted. The system is flexible enough to process different languages, layouts, and document types, including handwritten and stamped fields.

Discover logistics automation

For an international insurer, we realized a project that transformed the entire process of processing medical claims. The challenge was to quickly, accurately, and securely process large numbers of scanned claim forms from various insurance companies and in different formats.

Using an advanced platform, scanned claim documents are automatically recognized. Important fields such as policy data, insured quantities, amounts paid and outstanding claims, and breakdowns by type of care are directly extracted from the documents. The system adapts to different document structures and can quickly add new templates.

Optimize claim processing

For a regional government organization, we executed a project in which the processing of objection and court documents was fully automated. The challenge was to efficiently and error-free link incoming WOZ objections and court mail to the correct files, integrating various document types, barcodes, and databases.

Incoming objections are automatically retrieved from a secure SFTP location. The system recognizes the document type, reads assessment and file numbers, and links them directly to the correct records in the database. Court documents are recognized based on unique features and matched with existing files.

Discover government automation