When we talk about Intelligent Process Automation (IPA), we’re typically referring to the methods in which the Ephesoft Transact can augment Robotic Process Automation (RPA) systems.

There are two main use cases where intelligent capture can enhance business processes: (1) as a mechanism to identify and extract actionable business information that is used to initiate a business process, providing the initial data payload, and (2) and as document intelligence tool to be accessed from within a process via API calls.



However, a third and rarely discussed use case is the relationship between intelligent capture and workflow automation is the case where an RPA system extends the capabilities of Ephesoft Transact’s document process. In this case, information or data that is not present on the document and records analyzed by Ephesoft Transact could be pulled from external sources and then incorporated into the workflow.

Many RPA systems, such as UiPath or BluePrism, have rich user interfaces that provide end users with a low code option for building process automation workflows. This type of tool allows organizations to replace complex or manually-driven processes that would typically be accomplished by a worker. Take, for example, background research or information validation as a part of application review.

In this use case, we can assign Ephesoft Transact the task of document identification and data gathering. Working in tandem, the RPA system performs due diligence on applicant submitted information. Let’s break down the steps of this workflow:

  1. An applicant submits their document package for company consideration via email.
  2. Ephesoft Transact monitors the specified inbox for new submissions and ingests the email attachments upon arrival.
  3. The documents are categorized by type, and key metadata is extracted through a combination of rules-based logic and supervised machine learning.
  4. In order to confirm or vet applicant-submitted data, Ephesoft Transact calls the RPA system to validate the information. For example, a robot could check a third-party website such as LinkedIn to verify that the job title listed in the application packet matches the position detailed online.
  5. Any data returned from the robotic process that does not match the data in the application packet will be flagged for review.

Comparing document information to structured data sources such as an external database is a simple task and easily configured (and automated) in a capture tool like Ephesoft Transact. However, making that same comparison with unstructured third-party data – like text embedded in a LinkedIn profile – is highly complex. Without the use of an RPA tool to augment this document-centric workflow, a human employee would be tasked with much of this laborious fact checking.

It is difficult to imagine the ability of an RPA system to perform without a document intelligence tool. The ways in which the artificial intelligence of robots can augment smart capture workflows will undoubtedly boost straight-through processing and minimize required human “touches” in document-centric operations.

Interested in learning more? Check out our demonstration video library on YouTube.