How NBFCs In India better decisions with AI and ML

Spread the love

In recent years, the number of NBFCs ( non-banking finance corporations) in India has been growing steadily. But at the same, these NBFCs in India continue to face many business challenges and risks. More than often, NBFCs encounter financial frauds, non-payments and loan defaults, forged documents and fake credit worthiness or scores, as well as identification and authentication of customers.

Overall these factors play a key role in the core decision-making aspect for loan approvals and credit disbursements. And that remains a very difficult task for NBFCs in India.

However, Arvog Finance, Avanse Financial Services and Clix Capital are among the many NBFCs in India that have invested in emerging technologies for improving decisioning in their organisations.

Leveraging AL and ML

These NBFCs are leveraging artificial intelligence (AI) and machine learning (ML) powered tools and solutions to enhance key business processes ranging from decision making to credit risk mitigation to automation and more.

For instance, Arvog Finance – an NBFC that offers credits for educational purposes, gold and digital microloans across India. It has been leveraging the AI and ML-backed system for credit and predictive risk monitoring and supporting its back-end loan processing.

Arvog outsourced this decision-making system from Washington, US-based firm Lokyata – an AI and ML specialist for financial products. It has implemented and integrated Lokyata’s decisioning system with the existing loan management system and APIs.

Customer-centric and technologically relevant

The new system deployment has helped Arvog to enhance its product and service offerings. It has transformed Arvog into customer-centric and technologically relevant.

“After deploying Lokyata, we realised, how processes can be streamlined and how it has drastically helped us to reduce the turnaround time,” says Priyank Kothari, Director – Arvog.

“Initially, most of the processes were still manual and used to take a lot of time for manual processing and verification of documents. And then decide whether the loan is to be approved or rejected,” Kothari explains the pre-deployment scenario in his organisation.

But for Arvog, the deployment of Lokyata has helped to change the scenario. Its efficiency levels have improved than in the past, while the TAT (turnaround time) levels have been reduced.

Reliance on AI and ML-backed system

While Arvog has continued to run its financial services business relying on the AL and ML back decision-making system. Kothari’s reliance on this new system is beyond decision-making.

“Obviously, the reliance on these models (AI and ML-based systems) has increased more than before. It has helped us in reducing a lot of frauds. Earlier, documents could be easily replicated and sent to us. But with these models, I think fraud risks are being mitigated,” emphasizes Kothari.

The system has helped Arvog in detecting fraud and fake documents. For example, individuals duplicated the information mentioned on documents like payslips but income levels or salaries were unusual and misrepresented.

So, going ahead, Kothari wants to deploy more AL and ML algorithm-based models in the organisation and gain its benefits.

Like Arvog, Avanse Financial Services offering domestic and international educational loans has deployed an AI-backed credit decisioning platform from US-based Scienaptic. With the deployment, Avanse Financial Services has automated its loan decisioning with seamless education financing for students.

Automation and better credit-decisioning

Scienaptic’s AI-backed credit decisioning platform has enabled Avanse to improve speed to market with automation and better decisions. And at the same time, help Avanse Finance while minimizing risks.

“This collaboration and implementation of AI-driven credit decisioning will enable us to administer credit more effectively and automate the loan decisioning to offer seamless education financing to the student fraternity,” says Amit Gainda, MD and CEO of Avanse Financial Services.

“It will further strengthen our technological advancement journey dedicated to providing a hassle-free experience to our stakeholder ecosystem,” adds Gainda.

Data and the role of AI and ML

While the NBFCs in India have started to adopt technology but sparingly for some functions and not across organisations. Their biggest challenge is traditional data and working with it.

These NBFCs deal with traditional data sets, which are a mix of unstructured and structured. Moreover, these sets may not be in a digital or standard format that computer systems could easily capture electronically and process.

Manual processes like document collection and data verification consume more time and it directly affects the decision-making and credit underwriting processes. This scenario is quite widespread across NBFCs in India.

Dealing with structured and unstructured data

“One of the initial challenges that we faced regarding data is the intellectual amalgamation of structured and unstructured data to derive valuable information. Extracting the required data was, at times, challenging,” says Avanse’s CEO Gainda.

“Over the years, we have created a data lake. Its a centralised repository, designed to store, process and secure large amounts of structured, semi-structured, and unstructured data,” adds Gainda. 

This data lake, according to Gainda enables users to gain a deeper understanding of business situations. “They have more context than ever before, allowing them to accelerate analytics experiments,” he says.

Generally, the traditional data provides an overview of customers’ income and spending. But not insights into their behavioural finance, credit risks and other parameters. For NBFCs, these parameters are vital in the analysis and assessment of customers. As it helps in deriving confidence scores, credit worthiness and approvals, predicting financial risks and more.

Limitations of traditional data

“Our customer segments, usually have a very thin line of credit history. So we only had access to traditional data with respect to CIBIL scores, bank statements and ITRs (income tax returns). And this basically gives a view of how customers are financially performing,” shares Kothari.

The limitations of traditional data compelled NBFC Arvog to invest in emerging technologies like AI and ML-backed systems.

“But, with the Lokyata’s help, we got to know the difference between behaviour data of customers, which is very important for us to know and analyse, versus the traditional data that we had before,” points out Kothari.

“Obviously with artificial intelligence, you understand the behaviour of customers. And that also gives a lot more insights on their ability to pay,” continues Kothari.

Algorithms, behaviour data and underwriting credits

Arvog Finance has more than 1000 sets of algorithms and data captured through the Lokyata system coupled with behavioural data. With all these data at their disposal, Arvog underwriters have deeper insights to evaluate customer applications and derive their confidence scores.

“I think this gives us a clear view between the traditional and behavioural data. Traditional data was very limited in analyzing the actual ability of the borrower to pay,” points out Kothari.

“But with access to more and more behavioural data with the help of AI, gives us more insights and better scoring tool and ability to analyze whether the borrower is credible to pay or not,” adds Kothari.

Like Kothari, Avanse’s MD and CEO Gainda points out that Scienaptic’s AI-driven platform analyse multiple data and information to underwrite credits effectively.

“This process enables us to evaluate more profiles at a faster TAT which further supports us in making instant and quick decisions, reducing the overall risk factor,” explains Gainda.

The AI-driven platform, Gainda says allows their customers to witness a hyper-personalized credit experience with the right credit under optimised terms.

Risk predictions and mitigations

The scope of these systems goes beyond just helping NBFCs improve their decisioning processes. These systems are also helping NBFCs mitigate their financial risks and improve future growth prospects.

For instance, Avanse has created an AI and ML-based platform leveraging data to deal with financial risk and payment defaults. According to Gainda,

“Collection is an extremely important part of our business success. Hence, we have created a platform that uses data to build perspective models to provide early warnings on customers who are most likely to default,” informs Avanse’s CEO Gainda.

Avanse’s AI and ML-powered Collection platform is an initiative toward a Bounce prediction model under risk analytics. “It has been created to analyze and identify the accounts which might enter into the overdue (OD) stage,” Gainda explains how the system helps in mitigating risks.

“This analysis will be conducted with the help of artificial intelligence and machine learning analytics. After the vulnerable accounts are identified, they are transferred to the next level. We reach out to them for reminders at regular intervals at this level,” adds Gainda.

Globally, the BFSI industry has been increasingly investing and relying on AI and ML-based software tools today.

In 2020 alone, the global market size of AI and advanced machine learning (ML) in BFSI size touched around $7.66 billion, according to a report. And the market size is estimated to reach $61.24 billion by 2030, growing at a CAGR of 23.1% for the forecast period 2021 to 2030.

Reforms and technology-led regulations

Unlike banks, insurance and financial services companies. NBFCs in India are slow to adopt the technology due to various factors and constraints. And that’s where India’s central bank and regulators are playing a key role in introducing new reforms and technology-led regulations.

The investments in emerging technologies by Arvog Finance, Avanse Financial Services and several other NBFCs in Inda are helping them move forward and adhere to the Reserve Bank of India’s new mandate issued in February this year.

As per this new mandate, RBI has directed NBFCs with 10 and more branches to adopt Core Banking Solution (CBS) on or before September 30, 2025. And NBFCs with fewer than 10 branches have been advised to implement ‘Core Financial Services Solution (CFSS).

Although, it’s not mandatory for these NBFCs. But it is for their benefit, according to RBI. With this new mandate, RBI aims to bring all the NBFCs in India into a revised regulatory framework. It’s similar to RBI’s CBS mandate to all nationalise and private banks some three decades ago.

Leave a Reply

Your email address will not be published. Required fields are marked *