Appropriate data is the basis of every AI/ML-based solution. Most of us are aware that banks and other financial institutions have more structured data than businesses in any other sector. Srajan Agarwal of Elets News Network (ENN) conducted a talk with Hardik Thaker, DVP – Digital, Analytics and BI, Aadhar Housing Finance Limited, to learn more about India’s readiness to adopt technologies like AI & ML.
How are financial institutions adapting to a customer base that is changing quickly?
The central government’s numerous initiatives have led to a dramatic increase in digital adoption in India over the past ten years. Everyone is aware of UPI’s pre- and post-pandemic contributions to India. UPI transactions in the nation were 10.7 lakh Cr (10,72,792.68 Cr) in August 22 compared to 2.9 lakh Cr (2,98,307.61 Cr) in August 20, a 259% increase in just two years.
Customers from all demographic groups, including professionals, non-professionals, HNIs, and people with low incomes, have dramatically expanded the adoption of digital banking in T3 to T6 town. The majority of Indian financial institutions have so far completed their migration to Digital 1.0. However, such a significant shift in the population necessitates reevaluating current company plans. Neo banks and other fintech businesses are also making a contribution by innovating in order to better serve and comprehend the consumer. In the next days, we’ll see a relationship between big banks and these fintech companies.
On top of the current infrastructure, financial institutions are developing or collaborating with important partners to add layers of alternative data sources, automatic AI/ML services, process automation employing cutting-edge technology, and GIS services. They will be able to strengthen GTM, marketing, acquisition, and brand proposition with the aid of these layers.
The banking industry will become very commodity-based during the next five years. Think of electric versus gasoline-powered cars, for instance. Every automaker has a niche in which they specialise in producing gasoline engines, giving them a competitive edge. However, a battery cell is used in electric mobility. Compared to gasoline-powered cars, EV construction is simpler (or any other EV vehicle). Customer experience will be the only differentiator! The best example of UPI is PhonePe, Google Pay, and others. The Indian government is working hard to standardise banking and financial services. The organisation that best understands its clients and offers the finest customer service will succeed in the coming years.
Blockchain, AI, machine learning, the Internet of Things (IoT), SaaS, and other technologies are having a significant impact in a number of fields. How do you think these new technologies will create the foundation for future development?
In contrast to Bitcoin (or any other alt coin), blockchain is still in its infancy, thus I would recommend putting it aside. Blockchain technology is being used by many private sector businesses in the banking and IT sectors to produce solutions, but computation remains its major obstacle.
See, Arthur Samuel coined the phrase “the branch of study that offers computers the ability to learn without explicitly being taught” in the 1950s to describe the first-ever ML concept. Many academics tried to develop effective ML programmes throughout the course of the following six decades, but it finally took off in the year 2000 when we overcame ML’s greatest obstacle, “Computation,” and the rest is history. This time, I’m confident the solution won’t take six decades.
Also Read | AI and ML are altering the banking experience for customers
We can now examine how advances in cloud computing are to blame for the widespread adoption of SaaS, AI/ML, and IoT. With its vast array of tools and lack of concern for infrastructure scaling and maintenance, AWS was a pioneer in this field. As I previously stated, the market will be dominated by the businesses who best comprehend and service their clients. SaaS applications give businesses a platform to quickly reach millions of customers. The extended arms of these SaaS systems are thought to be AI/ML and IoT. Together, they create an ecosystem that allows customer-centric business models to converge.
How ready do you think India is to adopt financial technology like artificial intelligence (AI) and machine learning (ML)?
Appropriate data is the basis of every AI/ML-based solution. Most of us are aware that banks and other financial institutions have more structured data than businesses in any other sector. Although this is somewhat true, an institution cannot do more with the current organised data. Customers also object to financial service providers routinely scanning their accounts since it invades their privacy. Financial institutions are also developing alternative data solutions, either in conjunction with partners or directly from government organisations, such as NDSL, Perfios, GST APIs, etc.
The financial sector is enormous, but from the perspective of credit and lending, many organisations have tried automating credit underwriting using AI and machine learning models on customer data that primarily consists of financials (either personal or business), bank statements, and credit bureau records to understand RTR. Collaterally valuation is a challenging aspect, thus not everyone has achieved so far. To establish smooth data exchange procedures among financial service providers using AA (Account Aggregator) frameworks, there are numerous developing service providers and frameworks created by the Central Government.
Advanced AI/ML approaches are being used by lending organisations to increase productivity and lower costs for client acquisition and collection. In my opinion, the period has just begun, and additional use cases will emerge in the near future.
How will financial services be impacted by the rapid growth of cloud-based ERP, automation, and cognitive innovation?
Let’s start by examining the development of the banking industry. Only skilled individuals used the bank’s few central computers in the 1990s to provide basic banking functions. Later, banks in India started installing just one mainframe machine per branch, and all of the branches were now linked to the bank servers. Over time, it changed to include many PCs and users operating through branches; currently, numerous employees are concurrently connecting to a bank server. The net banking age of the 2000s begun. Customers can now use fiXed locations to access bank servers. From 2015 onward, users may access bank servers from anywhere in the world thanks to the advent of mobile banking. In the year 2022, “connected apps” are a development in mobile banking, allowing clients to access bank servers through third-party service providers using APIs.
One must realise that these journeys demanded significant structural change from every provider of financial services, whether it be CRM systems or Core Banking systems. Therefore, selecting the appropriate ERP solutions becomes essential for every organization’s success. With linear solutions, one cannot solve linear problems. (Increasing supply shouldn’t be done to meet rising demand.) The truth is that today’s consumer expects to be handled with special consideration yet lacks patience. Here, cognitive intelligence integrated into the heart of any good or service helps lower the financial service provider’s overhead expenses associated with providing client support.
Traditional IVR systems use fixed options and instructions and behave more like machines than people. Modern chatbots and voice bots are cognitively intelligent and can mimic human behaviour and thought processes to tackle complicated problems. The customer support channels for Zomato and Swiggy are the best known examples.
Any financial service provider can lower operational expenses by developing the appropriate solution, which will also enhance the entire customer experience.
How poorly do data analytics and AI support infrastructure development in the housing finance department?
As I’ve already indicated, the correct data strategy is essential for the success of any analytics or artificial intelligence programme. There won’t be an analytics strategy in 2022; in the digital age, strategy will suffice.
Let’s first examine the basis on which HFCs lend money to their clients. Initial and crucial collateral appraisal Capacity for repayment of the second customer. Very easy! Right. It’s not, in fact. Because collateral valuation is still quite subjective in India. I’m not referring to carefully planned residential or commercial developments in Tier 1 or Tier 2 cities.
At Aadhar, the majority of our clientele are low-income residents of T2 to T5 towns. You must now discuss the difficulties they had when having their loan applications reviewed. Although it has only recently started, the central government is pushing for digital land records, deed records, etc. to offer openness. Digital and data analytics adoptions have so far been bolstered by API-fication data from numerous government agencies, corporate, and public sector organisations.
