Share This Page

ShareShareShareShare
Author: Renukaradhya Channabasavaiah |06/24/19

Transformation in Lending with AI/ML Powered digitalization

Lending is one of the major revenue streams for banks and financial institutions (FIs) and that is the reason they are always looking at new technologies to innovate the function. In today’s banking generation, customer aren’t inclined to opt for the physical mode of loan requests’ submission and even banks expect customer to onboard via omni-channel platforms (Netbanking, mobile) for any kind of loan servicing - right from initial request till loan closure. Yet, banks are compelled to create loan/collateral documentation while processing the loan applications for audit and compliance purpose.

Processing a loan application from the sourcing agent until the loan disbursement process, involves many complex processes and workflows. These include customer onboarding, credit decisioning (internal / external), AML & Dedupe checks, fraud checks, collateral creation and linkage, document approval and final disbursement of loan to customer. Each process is inter linked and therefore dependent on the other to some extent.

Transforming lending for the digital age

As banks and FIs face competition from incumbents as well as entrants and struggle to meet the demands of today’s customers, the pressure to reduce TAT and implement latest technologies to automate lending has never been more real. Thankfully, by its very nature, lending is a function best suited for application of Artificial Intelligence (AI) and Machine Learning (ML) technologies as the main premise of lending is about handling humongous amounts of data to accurately assess credit worthiness. AI and ML based systems can also aid decision-making for every workflow predefined by the bank to completely automate the processes. Further, these systems can be configured to capture loan documents - right from the initiation stage till the disbursal stage to the document management system (DMS).

Forward-looking lending companies are also leveraging AI to give their customers a delightful lending experience by providing personalized and configurable journeys for mortgages, Home Equity Loans and Home Equity Lines of Credit (HELOCs), and home equity installment loans through a suite of user-friendly portals and mobile companion apps.

Five key AI use cases in lending

Application areas in the lending process where AI can play a significant role include:

  • Classification of documents

As lending involves lots of documents to be submitted and reviewed, the first step is to classify what document is submitted by the borrower to the system. Here AI can help find the unique attributes in the documents and ensures that the system identifies what document is ingested to the system from the borrower. Thereafter, it captures the data from the document to create a digital loan file.

  • Extraction of Documents

The next step is to leverage AI to extract the required details from the ingested documents by specifying what needs to be captured and where the data should be saved for further calculation of the borrowers’ front end and back end ratios. These ratios can predict the loan product suitable and available in the market that matches the borrower’s current income credentials. The goal is to meet the borrower’s requirements without the loan becoming a burden for both applicants and lenders in future.

  • Validation of documents

To validate the data extracted from the classified documents, AI helps ensure that all the documents ingested are consistent with the borrower’s details and with the required set of documents available for the borrower’s credibility calculations. Thereafter, ML algorithms can combine these credibility calculations with process automation techniques to identify inconsistencies and automate rule triggering to create decision ready digital loan files. Once prepared, AI can perform a number of underwriting functions in these files and can even spur reshoring of repetitive service tasks for lending companies.

  • Lending credit without having credit history

There are many such consumers who do not have recent credit or may not have credit at all. In such cases, ML based risk assessments can be conducted and can predict accurate credit scores to create credit profiles of such consumers.

  • Integrated Lending ecosystem

AI & ML can be used to seamlessly connect with internal and external systems of the larger lending ecosystem through RESTful APIs. AI can simplify the lending journey with secure and speedy loan processing and can enable innovations that can produce fundamental changes for both lenders and borrowers over the course of years.

Tread with caution

While these cutting-edge technologies hold immense potential for the banking and financial services industry, they should not be adopted vacuously. Banks should take a meticulous approach to first streamline the business process and then qualify the areas where these digital technologies can add value. Doing the right homework, before taking the plunge to transform lending is a critical imperative; else it would only lead to more complexities and adversely affect the business.

READ

Let's Talk About Your Needs

Thank you for your submission. We'll be in touch.