IDP Features

Component: DocuWare IDP

DocuWare Intelligent Document Processing (IDP) seamlessly extends existing document management solutions with AI-powered features. This enables customers to achieve a higher degree of automation in their office processes, increase efficiency and reduce operating costs.

Features

  • Splitting: Even without separator pages or barcode stickers, automatically divide batch scans into separate documents and ensure consistent further processing of the individual PDFs in company-wide workflows.

  • Classification: Classification assigns documents to different categories or groups based on their content, properties or characteristics. The respective workflows for each document type can then be started right away.

  • Extraction: IDP accurately extracts the important data from your documents, even in difficult cases such as complex line items and tables, as well as scans of crumpled or stained documents.

  • Extraction for Handwriting: Extraction using HTR (Handwritten Text Recognition) recognizes handwritten information in scanned documents, which can then be used as index entries for document routing.

Applications

  • Accounts Payable Process: Increase the level of automation by automatically extracting complex data from invoices and possibly crumpled delivery notes. Learn more

  • Mailroom: Fully digitize and automate your entire mail inbox, whether paper or email based. Learn more

Benefits

  • Reliable and consistent document classification ensures faster processing.

  • High degree of accuracy and streamlined automation regardless of the quality of the scanned paper documents.

  • Data from complex tables on delivery notes, etc. can be used immediately for further document processing.

Here's how

DocuWare IDP can be used in combination with both DocuWare Cloud and a DocuWare on-premises system. Various IDP packages are available to meet the needs of IDP functions.

There are several ways to integrate IDP functions into DocuWare:

  • via Workflow Manager: classification, extraction incl. handwriting

  • via StorageRobot: splitting, classification, extraction incl. handwriting (but no extraction of individual items for transfer to DocuWare table fields)

  • via Connect to Mail and IDP configuration: classification, extraction incl. handwriting

You also have the choice of which type of AI model you want to start with:

  • Pre-Built: Models come pre-configured with established structures, defining the data fields they extract or the classes they classify.

  • Fine-Tune Pre-Built: If the pre-built model performance is not optimal for your specific data, you can fine-tune the model by providing it with more specific data.

  • Build Your Own Model: You can provide the data for your unique case to enhance performance where a pre-built model might not yet meet the standards for specific document process.