Content Management and Capture

 View Only
  • 1.  IBM Document Processing Extension (DPE)

    Posted Wed June 05, 2024 02:42 PM

    Introducing IBM's Document Processing Extension (DPE), a powerful solution that harnesses the capabilities of AI to automate document handling tasks seamlessly. Utilizing low-code application capabilities, DPE allows users to design sophisticated deep learning models without the requirement of data scientists. Trained models can be easily applied to recognize various document types, including invoices, bills, purchase orders, tax forms, and more. This flexibility enables the creation of intelligent document processes and the extension of applications to leverage uncovered patterns and insights. 

     DPE services are strategically designed to instill high confidence levels, minimizing the need for manual validation, and focusing on the automation of data extraction. By categorizing and extracting documents or batches of documents, DPE ensures swift identification and correction of issues.

     What's Different:

    •      DPE is designed to provide similar document processing functionality as IBM's Automation Document Processing (ADP), which is available in the Cloud Pak for Business Automation (CP4BA) but as a standalone purchasable product.

     •      DPE boasts a substantially smaller footprint than ADP. This efficiency enables cost savings without compromising performance.

    •      DPE is the evolution of document processing, offering enhanced efficiency, automation, and adaptability.

    •      DPE takes advantage of Docker for its runtime environment, offering a lightweight and cost-effective implementation. This choice ensures a seamless and efficient deployment process for organizations looking to optimize resource utilization. There is no dependency on OpenShift.

     •      Better with Datacap - DPE and IBM Datacap work synergistically to form a complete IDP solution.

    How is DPE Different from ADP?

    DPE is functionally equivalent to ADP in the context of document processing performance. DPE is designed to be a lightweight version of ADP which does not require the OpenShift prerequisite. DPE requires a minimum support configuration of 6 cores which is a substantially smaller footprint than APD.  DPE does not a have runtime user interface (UI) for user verification and therefore is not a standalone solution; it must be a part of a larger solution (typically IBM Datacap).



    ------------------------------
    Krish Lakshminarayanan
    IBM Global Tech Sales Leader for Capture & CMOD
    ------------------------------


  • 2.  RE: IBM Document Processing Extension (DPE)

    Posted 30 days ago
    Edited by Olivier Baltus 30 days ago

    Proposing DPE as an alternative to ADP is a brilliant idea. It's much simpler to deploy, and there is clearly a real need in terms of classification and extraction as a service.

    In my company, we're deploying  "DigitalMail Room" project which combines Datacap with ADP, and we would love to replace ADP with DPE which in this case perfectly meets the needs. No need to have a CPE, we only use the REST APIs.

    The problem is the change in approach to licensing costs. With ADP, the customer can process an unlimited number of documents, unlike other solutions on the market;  which is a key differentiator in the pre-sales process. But,unfortunately, the DPE licensing method is different and involves purchasing packages of processing pages in advance, which immediately becomes less attractive to customer, especially in the case of processing incoming mail.

    I really hope that IBM changes its mind and offers the same licensing method as ADP.



    ------------------------------
    Olivier Baltus
    NSI Luxembourg
    ------------------------------