Improve throughput by 3X and increase the invoice straight through processing to over 90%
A leading telecommunications service provider attempted to implement an intelligent data extraction solution to process invoices received from multiple telecom’s (AT&T, Verizon, Century Link, Frontier) using a leading intelligent data processing (IDP) platform, to increase the throughput and accuracy of its invoice processing function.
The problem is that the IDP platform did not work well with complex documents: multi-page, unstructured invoices in up to 44 different formats; data that needed to be extracted from free-form text and tables; and data elements on different pages within the same invoice format.
Without the automation solution, the service provider would continue to incur significant penalties and the potential loss of business from impacted clients.
The Solution
The MOZAIQ services team first implemented a document indexing platform. By classifying the multi-page, unstructured invoices into discrete pages, the data could more easily and accurately be extracted from floating tables and fields.
The data output was further processed using custom python scripts, enabling greater accuracy and allowing higher straight through processing.
An RPA BOT was programmed to consolidate the extracted data and to then upload the data into the SAP system of record. This was supported by a human in the loop process that was integrated into the flow to process any exceptions prior to uploading the data.
The Benefits
The combination of the Indexing Platform, the Custom Python Scripts for pre- and post-processing, and the optimization of the IDP platform enabled the client to improve their invoice processing throughput by 3X, increase the straight through processing of invoices to over 90%, and to increase the accuracy of the data prior to uploading into the system of record to 99%.
The automation solution reduced the late payment penalties to near zero, and raised the level of their customer service, reducing the risk that their clients would leave them for an alternative service provider.