How banking industry transforms the turnover of documents

Nikita Kalinin - 27th May 2022

Intelligent documents processing (IDP) : How banking industry transforms the turnover of documents


Today banking industry is being shaped at a very fast pace: due to the COVID-19 pandemic, the industry faced a decline, which, however, was still less harming than the financial crisis of 2008 (ROE in 2020 – 6.7% while ROE in 2008 – 4.9%). Another issue is that currently around 51% of banks globally operate with return on equity (ROE) below their cost of equity (COE). So as not to be depended only on factors which are beyond banks’ control (liquidity, support from the government, interest and inflation rates), banks shall step on the path of digitalization and accelerate on hyperautomation, which is a structured approach used by companies to implement automation for every function, where it is applicable. Another reason for moving towards hyperautomation is the increase in the revenue stream driven by originate-to-distribute model (from 45% in 2014 to 55% in 2021).


IDP implications

Intelligent document processing (IDP) is playing a significant role in processes’ automation within banking industry. It shall be outlined that IDP may help a banking corporation from several perspectives: firstly, it helps to reduce costs of processing large number of documents such as mortgage forms, credit applications and other bank documents by removing manual input part of the process. Hence, we can highlight the second advantage which is a significant decrease in time needed for data input of any type of banking documents. It is estimated, that on average the process of data input goes 5 times faster with IDP than if it is handled manually. Additionally, due to vast volumes of data, which is to be uploaded manually, it is not unusual when a bank employee makes a typing mistake. IDP helps to minimize the risk of any kind of inaccuracy errors, that are linked to human-factor based mistakes.

Using our easy to use annotation and dataset creation platform users can annotate their document data easily and accurately.


Assessing IDP impact

It is not easy to accurately estimate the impact of IDP adoption within banking industry because it has been only a few years that a list of banks launched implementation processes of IDP solutions. There are enough use cases, however, to acknowledge a general positive impact of IDP on banks’ operations. Thus, the majority of observed banking/mortgage companies state the IDP’s ability to extract data with high accuracy as a major reason to implement this solution. Smart Layers has a capacity not only to extract data from different types of documents, but also provides high level of accuracy (e.g. 98%). The platform can also assess the IDP impact on invoicing: on average, around 25% of invoices include errors of different kind. Smart Layers provides a “pre-trained model” fully dedicated to invoices. Another case, where IDP could be used, is a credit/mortgage document processing. In case if a document from the required package is not compliant with the stated standards, Smart Layers’ classification models would be able to identify the inaccuracy and request an applicant to submit a proper version of the document, outlining a problem with the initial one.