Ephesoft Accounts Payable Solution for Alfresco, Documentum and Hadoop

Based on multiple recent client efforts, TSG is formalizing our solution for Accounts Payable.  This post will present how Ephesoft Smart Capture, a open source document capture solution, can be combined with OpenMigrate and HPI to deliver a more robust invoice and check processing solution for Alfresco, Documentum or Hadoop

Accounts Payable – Background

Many of the initial founders of TSG worked for Andersen Consulting back in the mid-90’s and met on an Accounts Receivable factoring (much like accounts payable) image processing solution for a client that was to later become part of GE Financial Services.  One of the most difficult jobs was setting up a template for each vendor (based on an example image scan) that would identify where certain key fields were located.  For GE Financial Services, we had to set up one template per vendor (600 templates and counting) and the templates would often need to be updated for a variety of reasons (paper change, scanner change, font change….).  Once all the templates were created:

  • Invoices would be arranged with separator sheets to identify which documents are invoices and which are attachments.
  • Invoices would be scanned
  • Recognition software would identify which template should be used for the OCR capture of different fields.
  • The template would extract the fields based on the template
  • A user would review the results based on confidence requirements for input into the image management and potentially accounts payable system.

Typically fields captured would include invoice number, purchase order number, invoice date, payment terms, vendor address, invoice amount and other fields based on invoice requirements.

Ephesoft Smart Capture– A more intelligent, “templateless” approach

While many of the capture technology vendors have evolved, the core components remain similar.  Ephesoft, as a new vendor, leverages a modern approach.  With Smart Capture, Ephesoft utilizes powerful OCR algorithms to extract information without templates. With Ephesoft:

  • Invoices and attachments are scanned or electronically captured from email/shared drive or an external system.
  • Ephesoft OCR’s the entire document looking for key words that distinguish the document as an invoice versus other types of documents.
  • Ephesoft extracts the key fields based on their values and proximity to other key fields. For example, the Invoice Date might be located to the immediate right or under the words “INVOICE DATE”.  Attachments are recognized as attachments without needing separator sheets.  Ephesoft doesn’t rely on templates or location of the fields on different documents.
  • A user reviews the results based on confidence levels.

Ephesoft combines other modern features including the ability to ingest PDF or other documents electronically (without scanning), complete browser based solutions, and validation using database lookups.

OpenMigrate and HPI – Completing the Solution

Once data and images are captured, the TSG Accounts Payable solution leverages OpenMigrate to ingest the invoice images into Alfresco, Documentum or Hadoop.  Once in the repository, our High Performance Interface (HPI) allows viewing of the invoices as well as other critical documents based on the document object model.  For example:

  • Purchase Orders – If captured on the invoice, HPI can relate the original purchase order to the invoices that have been processed against it.
  • Checks – If captured on the check/remittance, HPI can relate the invoice to the check that paid for the invoice.

HPI provides other key ease of use capabilities for accounts payable including:

See a video below of our solution with Ephesoft in action.


Ephesoft, combined with TSG’s OpenMigrate and HPI, can substantially automate accounts payable with a simplified capture solution along with a robust interface for processing of invoices and other related documents.

Let us know your thoughts below: