Lendingkart bets on AI for smart loans, IT News, ET CIO

By Dhrumil Dhakan

With $ 1 billion in loans disbursed at around 1.3 lakh of MSMEs across India since its inception, Lendingkart has experienced rapid growth in MSME financing and one of the main reasons for this is its platform of credit information.

Lendingkart’s proprietary underwriting model has been instrumental in providing 250,000 loan sanctioned loans over the past 6 years, but it is important to understand how it is used by the company itself. .

“These are MSMEs who have a small Kirana store, an electronics store, a coaching course, a tea room, etc. who need working capital to grow their business. The majority of these MSMEs have a slim credit history and have not been introduced to the formal credit economy. In order for us to expand our reach and provide MSMEs with working capital funding, we use alternative data and a proprietary ML to understand our clients’ creditworthiness and don’t just rely on office scores. LendingKart has optimized its underwriting algorithms over the past 7 years, with a strong track record, ”said Manish Bhatia, President – Technology, Analytics, Capabilities, Lendingkart speaking of Cred8, Lendingkart’s proprietary underwriting platform who decides in seconds whether the client is suitable for a loan.

Alternative data involves details like bank transactional information, customer address, company details, census, etc. Manish explained in more detail the evolution of the model saying, “We have been very nimble in introducing additional alternative data sources, and each year we update our model majorly. Currently we are running on version 8 and we are moving towards version 9 of Cred8. We started with a rules-based subscription with an assisted manual review, as the complexities increased we started using Random Forest and currently our model is built on XG Boost ML technology with around 80% of approvals being processed. automatically. What is the best in the industry in our segment today. “

“Using the Cred8 platform, we are able to segment clients into risk categories based on their alternative data, which has allowed us to extend access to credit even to MSMEs with ‘end clients’. , that is, new to credit, ”Bhatia added.

Since data is essential to the credit loan process, its security is of the utmost importance to the business and, in particular, to the customers themselves. Since Lendingkart was born in the cloud, the company relies on a zero trust policy to ensure data security, assuming every call to it is untrustworthy. “Personally Identifiable Information (PII) data is segregated for maximum security using the Zero Trust architecture,” Bhatia says on how it ensures data security.

With no retail presence, the entire LendingKart process is handled online. MSME customers discover LendingKart through different channels such as organic search, marketing, aggregation partnerships and finally through our DSA Offline to Online (O2O) portal.

“We have created a distribution platform, so the first two which are digital are also included in Xlr8 but the most powerful piece where we have DSAs, that is to say feet in the street, basically financial agents and MSMEs is a segment where not everyone is going to be comfortable doing it online and very similar to the democratization of credit, we have basically created this platform where DSAs can integrate our platform as an agent and we’ve created a very smooth portal for them to use, work with their clients, enter their file on our platform, and once it’s entered it’s fully digital, ”Bhatia explained. .

Bhatia believes that Xlr8 has also been able to make online transactions more transparent. Lendingkart has also built an open architecture to ensure that banks and NBFCs can integrate and use our E2E 2gthr lending platform. “NBFCs and banks want to be part of this ecosystem because they want to expand their reach and disburse loans on their books, so with the 2gthr end-to-end platform that we created – Xlr8 gives access to PAN India online and O2O Client (offline to online), Cred8 provides state-of-the-art subscription, ZeroTouch offers the best customer experience up to disbursement and Collec10, our collection platform, gives them the ability to recover in the event of default Bhatia said.

This E2E platform allows banks to access customers passing through our ecosystem funnel.

The lending company is also currently working on an end-to-end collection platform named Collec10, which is already in use internally, which enables Lendingkart to track the performance of collectors, in order to maintain the quality of collectors.

“We also use machine learning to understand how and when to pay, what kind of offers to offer customers and how to reach them. So this is all completely digital now. Thus, all NBFCs and banks on board 2ghtr already use Collec10, ”he said.


About Mary Moser

Check Also

How Buffett is posing as a crook on rising interest rates

Warren Buffett’s Berkshire Hathaway benefits from low-cost investment funding – Image: Shutterstock Warren Buffett’s Berkshire …