How Machine Learning Can Redefine Lending

Nowadays, financial lending organisations are finding ways to integrate their customers quickly and securely to provide them with all their products and services through Digital Onboarding.

People are cautious about taking a loan online. But, both lenders and their consumers are unaware of how the digital loan onboarding process is creating a revolution in financial services through an ease of approach and seamless journey, even in remote places. They not only provide an all-in-one platform starting from Loan Origination to loan recovery but also save a lot of time and hassle.

CloudBankin’ Co-Founder and Director, Mani Parthasarathy, and Domain Consultant of CloudBankin, Rajeshware Srinivasan, have come together to give an in-depth outlook on how Loan Origination is done through Digital Onboarding. So, lend your ears to hear all about it in the episode or devour our conversation below to learn about it.

Mani Parthasarathy

Hello, Rajeshware. Thank you so much for your valuable time. Can you start by giving your viewpoints on what forms the basis of the whole customer onboarding journey?

Rajeshware Srinivasan

Two factors are considered while forming the basis of the complete customer onboarding journey.

1) The first is to understand and identify customers’ paint points to set clear goals and the type of loan product the customers require.

2) The second is to perceive the ticket size for a loan and the terms that the NBFCs provide.

Mani Parthasarathy

Please brief us on how the customer journey differs for various onboarding flows.

Rajeshware Srinivasan

Three scenarios define the customer onboarding journey.

1) An urban population with technologically profound borrowers goes with the Do-It-Yourself flow.

2) The rural population depends solely on loan officers on behalf of borrowers to aid in the onboarding.

3) The semi-urban and semi-rural population has an assisted flow for the mix of urban and rural borrowers.

Going with all three flows is generally regarded as an excellent course of action.

NBFC wants to bring out a Unique Selling Proposition for their customers. This causes variations in onboarding a customer journey.

Mani Parthasarathy

What are the prerequisites required to strategize the Digital Onboarding journey for a loan product flow of NBFC?

Rajeshware Srinivasan

NBFC lays out its credit policy for the loan product to be launched. From India’s perspective,

1) KYC is the first stage where the collection of the customer’s data such as name, address, industry, employment type and other data points are done. For partnering NBFCs, the KYC is also applied in this stage.

2) Then, the customer’s financial worthiness is checked when income details are provided.

These are the mandatory stages for underwriting a loan.

Mani Parthasarathy

What would be the cost per loan application for a traditional NBFC? 

Rajeshware Srinivasan
It is ideal to attach only the direct cost to a loan application since the indirect cost is quite challenging and complicated. For instance, if an NBFC serves one particular area, the cost is for that serviceable area alone. Hence, based on that, anywhere above Rs. 500 becomes the base price for a traditional onboarding journey for a customer. In the case of Digital Onboarding, it is much less, which is anywhere between Rs. 100 to Rs. 150, when the direct cost is added for the API being integrated into a loan application for KYC. Depending on the number of steps involved in the onboarding process as defined by various NBFCs, the cost will either increase or decrease.
Mani Parthasarathy

API has become a driving force now for every process. The most used APIs in India are Aadhar and PAN. Rajeshware, what are the challenges you faced in this regard?

Rajeshware Srinivasan

PAN API is much easier when compared to Aadhar API, which includes the customer’s name, PAN number and mobile number for verification.

Even though the Aadhar API is made available, it often causes friction in a customer’s journey because it involves OTP authentication where it is forwarded to the registered mobile number. This takes the customer to the UIDAI site to download the XML file of their Aadhar and then upload it into the system. It is quite a headache. Therefore, it is essential to go for biometrics or online verification in such cases. But, online verification is unstable. And its restoration of services takes time. Thus, NBFCs can face problems in accessing the document for verification.

Mani Parthasarathy

What is your stance on C-KYC?

Rajeshware Srinivasan

C-KYC is available today for banks, financial institutions and individuals to complete the KYC process. A probable flow could be after verifying PAN using PAN API. C-KYC can be integrated followed by a face match to complete the KYC process. This is another alternative available to Aadhar API.

Mani Parthasarathy

For a seamless onboarding process, what is the list of APIs required to be integrated?

Rajeshware Srinivasan

Aadhar, PAN and DV are available for most KYC processes. Additional documents such as GST, IT returns and bank statements can also be added.

Mani Parthasarathy

Can you explain the customer due diligence process flow?

Rajeshware Srinivasan

The Customer Due Diligence process involves two stages.

1) In the KYC stage, inter-matching of documents is done for verification. NBFC handles KYC and after that, the loan application moves to the underwriting stage.

2) If the documents collected in the first stage are not enough, then we go collect additional documents.

Mani Parthasarathy

Let’s take an example. A person has uploaded his Aadhar with his image on it. That same person uploads a live video of himself for verification. So, is there a system available to match that Aadhar image with his video image?

Rajeshware Srinivasan

Yes, it is available. Financial institutions that use video authentication for their KYC do this in four ways.

1) An individual from the institution is assisting a customer to enter into a video panel and verify the documents simultaneously.

2) A liveliness check is performed where the customer is asked to do an action like face movement or blinking eyes. The captured image is then matched with the documents he provided.

3) The customer can also go for a selfie image verification.

6) Geo-tagging is added to know whether the customer is located in India or not to sanction the loan amount to him. RBI has said that the image of the customer with his current coordinate is to be imprinted on any loan management software.

Mani Parthasarathy

How are frauds checked in the Loan Origination process?

Rajeshware Srinivasan

The four ways of KYC as mentioned earlier are Aadhar, PAN, video and face authentication, and geo-tagging. The KYC process should go through all four phases to prevent fraud. If KYC is incomplete, then fraud happens.

Mani Parthasarathy

In the recent case with actress Sunny Leone where her PAN number was stolen to attempt to get a loan in her name, how can we avoid such a breach against hackers?

Rajeshware Srinivasan
It is possible for anyone to access an individual’s PAN and Aadhar details, provided the hacker has that individual’s mobile number and other information in hand. Even two-factor authentication is just a minimal step. It does not define or say whether the person accessing the authentication code using the mobile number is a valid user. That’s where facial authentication or video verification comes in to prevent any data breach and ensure the person is who they say they are.
Mani Parthasarathy

How does a face match & name match work in the API?

Rajeshware Srinivasan

Today’s API algorithms are getting intelligent to generate any information and authenticate an individual’s face and name match. The facial structure, contour, depth, and dimension are considered for face matching, and the system generates an impression in terms of percentage, say about 75% or 80% match. The matching impression percentage depends from system to system. A similar thing is done for the name match. For biometrics, the system generates a point match instead, like an 8 point match or a 9 point match.

Mani Parthasarathy

What challenges do we face in OCR reading on the Digital Onboarding process?

Rajeshware Srinivasan

The challenge often presents itself in the cost portion, which involves two types.

1) The first is the conversion cost where documents are scanned and converted into digital data.

2) The second comes the verification cost where documents are checked for validation.

OCR reading is still not evenly centralised in all stages, creating issues. It is forced to move to manual mode. Thus, for a customer, it is not an ideal experience when they are told to upload additional documents or require them to be assisted by someone through the process. The current success rate is not very good and the cost is also not very welcoming when relating to that success rate.

Mani Parthasarathy

For NBFCs who deal with cash-flow lending for their customers, how does the bank statement analyser work?

Rajeshware Srinivasan

Bank statement analysis can be done easily when the customer’s data is extracted by accessing internet banking from the bank portal, and the time period for which the information is required. The API that analyses bank statements will do the mathematical calculations and give out the balance for the credit amount. A lot of data is extracted, but only the necessary ones are considered for the calculation by the analyser.

Mani Parthasarathy

Like the bank statement analyser, how does the SMS analyser work?

Rajeshware Srinivasan
Mostly with respect to individual customers, cash flows such as
Credits Withdrawals
Cheque deposits Cheques
Salary ATM cash withdrawals
Rent Online Transactions and Wallet Transactions
Any other deposits to accounts EMI payments
Online deposits received Credit Card Transactions
SMS analyser is able to analyse them almost accurately. In organisations, it is more streamlined.
Mani Parthasarathy

Rajeshwari, can you give an overall idea of what CloudBankin does?

Rajeshware Srinivasan

Cloudbankin is a system that handles right from the loan origination to the loan recovery. It is an API based application and very compatible with any system and it has a fully-fledged loan management system to support most loan products. A customer can go through all three onboarding flows. There is a mobile application and a web application available to suit any device. It is also integrated with other financial institutions for loan recovery. Thus, it is an end-to-end process that takes the customer on a complete lending journey.

Mani Parthasarathy

What are all the key operational risks available in Digital Lending?

Rajeshware Srinivasan

Today, API algorithms are upgrading frequently. Due to this, issues like PAN not working and Aadhar access being restricted for a few hours occur. These are the required risks a fintech should look into. Others are reasonably manageable. The advice generally given to fintech companies is that they should never build a complicated system. They need to keep it simple and take it step-by-step.

Mani Parthasarathy

Keeping all the data collected and the risks involved, how can CloudBankin quickly disperse a loan?

Rajeshware Srinivasan

CloudBankin has the potential to sanction a loan quickly. However, the time of entry into the app to disbursement depends on the KYC & underwriting process followed. The verifying data confirms that from the entry point into the app to disbursement of a loan can be under 10 minutes.

We hope you could grasp a lot of knowledge about how to expedite loan origination through digital onboarding, which the experts in the domain imparted. Don’t forget to stay tuned for more upcoming blogs from CloudBankin if you enjoy such content!

Frequently Asked Questions

What is loan origination system software?

A loan origination system software streamlines, automates and manages the entire loan process, from application to underwriting, approval, documentation, pricing, agreement, and disbursement. It benefits lenders by reducing errors, saving time, enhancing the customer experience, and improving the operational efficiency of their lending business.

Related Post

Power of Loan Origination System

From Paper to Paperless: The Power of Loan Origination Software

Overview In a world that is rapidly moving towards digital

How Digitization in Loan Origination Process is solving the challenges of Banks & NBFCs?
10 Stages in the Loan Origination Process

10 Stages in the Loan Origination Process

Introduction: Loan origination system (LOS) starts when the lending entity

© 2024 LightFi India Private Limited. All rights reserved.
(Formerly known as Habile Technologies)

×