Our goal is to enhance the UI and UX to significantly improve the customer experience: Vivek Dhavale, CTO, Equitas Small Finance Bank

In an exclusive interaction with Express Computer, Vivek Dhavale, Chief Technology Officer, Equitas Small Finance Bank, shares the bank’s key milestones in their digital transformation journey, particularly focusing on efforts to streamline customer onboarding through initiatives like eKYC and eSign. He highlights the bank’s success in achieving a high adoption rate of digital processes in their microfinance operations and discusses their strategic use of AI and machine learning to enhance fraud risk management and underwriting processes. Additionally, Dhavale touched upon the challenges and future roadmap for technology at Equitas, including plans for a super app to integrate various banking services, and the bank’s position in technology adoption compared to competitors.

Equitas Small FInance Bank has been a leader in providing financial services to unbanked segments in India. Can you walk us through the key milestones in your digital journey that have helped you achieve this?

So, we have undertaken multiple initiatives, primarily focused on streamlining customer onboarding. During the customer onboarding process, there is typically a significant amount of paperwork, involving typing and re-typing, and so forth. To address this, we introduced eKYC, which interfaces directly with UIDAI using biometric or OTP verification methods. Once eKYC is completed and KYC is confirmed, NSDL PAN validation is automatically processed. 

Furthermore, we have implemented eSign for agreements, sanction letters, and application forms. At different stages, eSign is triggered automatically. Customers receive an SMS prompting them to click on a link, where they can verify their Aadhaar details with OTP for eSign. This process eliminates the need for extensive paperwork, stamps, and physical documents. 

Additionally, bank details validation is facilitated through account statement aggregators like Perfios. We also ensure timely updates throughout the process. Essentially, our aim is to minimise paperwork wherever possible through these initiatives.

Are going completely paperless right now?

It can’t be entirely avoided because certain secured transactions, such as housing finance, require collateral. However, for all unsecured transactions, we are striving towards that goal. Specifically, on the microfinance side and in rural business, this approach has proven highly effective. Surprisingly, microfinance leads in the adoption of eKYC and eSign among all our business lines. We’ve observed that at least one member of each household typically possesses a smartphone.

In which specific pockets or clusters of the country is your microfinance more effective?

So, we operate mainly in tier 3 and tier 4 cities. The microfinance business is widespread across all villages, you know.

As a bank, our primary focus is on the southern regions. We also have a significant presence in Maharashtra and the northern areas. Approximately 40 to 50 percent of our operations are concentrated in Tamil Nadu, Karnataka, Andhra Pradesh, and Telangana. 

So, in that sense, it covers a broad spectrum. Nearly 96 to 98 percent adoption rate is seen in microfinance.

In contrast, vehicle and housing finance sectors have not seen as much traction. However, microfinance has seen remarkable adoption rates, spreading rapidly like wildfire.

And when did you start this digital transformation journey?

It’s been almost 18 months now that we’ve been consistently working on this. Microfinance saw the fastest growth. Initially, adoption was only around 50-60 percent. However, last month we reached 98 percent. Until last year, it hovered around 96 percent, which is quite encouraging. April and May also showed excellent adoption rates. I was pleasantly surprised and very happy about it. This can definitely be seen as a key milestone for us.

AI/ML are revolutionising the financial services industry. How is Equitas Small Finance Bank leveraging AI/ML to improve its operations and services?

Honestly, when discussing AI and ML and their use cases, there haven’t been many impressive examples so far. However, after some internal brainstorming, we decided our initial focus should be on fraud risk management in transaction processing. This is our primary target at the moment; although we’ve just begun, we’re not yet fully operational. Our goal is to ensure it’s not just a superficial AI/ML process. We aim to truly understand and improve the mechanisms of thinking and learning, which is fundamental to effective AI/ML.

Our second priority is underwriting. Currently, all credit approvals require manual review by a credit manager who evaluates scores. We believe a significant portion, around 60-70%, could effectively be handled by AI/ML. Additionally, we aim to incorporate any recurring deviations based on specific parameters into our processes.

Do you have specific parameters to gauge the success of your efforts so far?

The best part of transaction monitoring is that you can see how many false negatives we had earlier and how many we will have after this. Currently, we have around 70 to 72 percent false positives. In the first phase, we aim to bring it down to 60 percent, then to 50 percent, and eventually below 50 percent. It will be an evolution, a journey that we will continue.

Right now GenAI is the hottest topic in the industry, so, are you also leveraging its benefits?

We have started implementing this first in learning and development (L&D). For L&D, it is a little easier because we have set up documents. Practically, if everyone has to read them, they might not do it consistently. As part of their day-to-day tasks, if they have a query and want to ask something specific, we want to bring relevant information from across two or three documents. This will allow them to find the necessary information and complete their processes efficiently. If we can do that, a lot of queries from branches or compliance issues will be resolved. Often, these issues arise because someone read a document and forgot or misunderstood it. Hopefully, we’ll be able to address that.

On the customer service side, there are many interactions. Customers keep sending emails, making calls, sending SMS messages, etc. Based on that, we aim to consolidate their queries and provide a holistic answer. Additionally, we hope to draft good emails that accurately address their concerns. This should help increase customer satisfaction, even if only by a small percentage. That’s what we’re trying to do and it will definitely enhance customer satisfaction.

Are you going to launch chatbots as well for customer service?

Yes, so right now, this project will implement a chatbot for email communication, including consolidation and email replies. So far, we have tried to implement it in 11 languages to cover most of the country. We always start with English, then move to Hindi and Tamil, followed by other languages. We will implement all languages for sure.

Data security and privacy are crucial concerns for all the financial institutions. How is Equitas ensuring robust data protection practices in the digital age, especially considering the needs of the underbanked population who maybe less familiar with data privacy?

We are trying to take care of the data centre side. We have implemented many security tools, so most of the crucial customer demographic data is masked. This means nobody can see the exact PAN or other sensitive information unless they are legitimately logged into the application. With data masking, DLP, and other security measures, the data centre is secure for customers. 

There is no paperwork or other physical documents collected, even in xerox form. Even if data is masked, such as four-digit numbers, it will never be exposed because of eKYC. Even loan account numbers and other details are protected. We have ensured this security, plus on the collection side, we have measures in place. 

Subsequently, on a regular basis, whenever a customer repeats a transaction, the collections officer will generate a QR code. When the customer scans the QR code, they can see their account information, name, balance, and received money. This means there is no exchange of data, ensuring data privacy and security. No paper or account number gets exchanged, and this way, we are taking care of data security. Additionally, we are focusing on AI and security.

How can a balance be achieved between using AI to mitigate risks and addressing the enhanced risks that AI itself introduces, such as with generative AI?

In any implementation, we try to build controls and checks to monitor responses and actions, whether it’s a security issue or not. It can be a balance of using AI and security. This allows us to see if what is happening is in line, varying, or completely different from past occurrences. This culture is applied to normal transactions as well. 

For any activity, whether it’s RTGS, UPI, MPS, the switch, or CBS, we ensure everyone is in PCI compliance and synchronised. We maintain control and a reconciliation mechanism.

Similarly, we aim to build systems with random checks because compliance and audit require them. Additionally, we want to ensure there is relevance between the content of the customer’s email and the response email. While the risk may not be 100 percent mitigated, we can be confident that most aspects are taken care of.

This is the risk mitigation strategy we apply across all transactions, implementations, and even AI.

What are some of the biggest technological challenges facing small finance banks in India, particularly those focused on financial inclusion? How is Equitas addressing these challenges?

Challenges are there. I do not foresee or believe it will be a technological challenge. I think the challenge will always be in the execution part of it. We have started with AI/ML now, but the thought process began a quarter ago. The speed at which we want to progress is always a challenge. It is because of technology, the vendors we want to work with, and the expertise they have. Most of them may come with very simplistic use cases and require a big investment, which doesn’t make sense. 

We could rectify this only when we saw some concrete use cases, at least in phase two, if not in phase one. That is when we started work. Additionally, being a regulated entity, if any regulation compliance-related item comes up, any security breach, or any security-related aspect arises, they will always have a much higher priority than the execution of these projects. I think those are the challenges that actually hamper the execution part, not so much the technology or the usage part.

Looking ahead, what does the future roadmap for technology at Equitas? Are there any emerging technologies that you’re particularly excited about exploring?

We have undertaken a number of initiatives. The simplest ones include our internet and mobile banking. We want to enhance the user interface (UI) and user experience (UX) to significantly improve the customer experience. Our goal is to ensure that when a customer logs in, they immediately see what they want rather than a menu showing their Casa balance or other details. We aim to present recent transactions and frequent beneficiaries at the top.

We are on a journey to transform this experience into a super app. This means there will not be separate apps for onboarding and mobile banking; everything will be integrated. Whether it’s payments, onboarding, servicing, credit cards, FastTags, or prepaid services, it will all be within a single app. This project has already started and is expected to launch within the next month or two, or by the next quarter at the latest.

Internally, we have made significant improvements to our core systems. We have eliminated many legacy applications and migrated to new core banking systems. For the loan and asset business, we now have a proper loan origination system equipped with digital KYC (eKYC), e-signatures, and all necessary validations, except for the unavoidable physical documents. 

Moving forward, we aim to consolidate our platforms so that onboarding, whether for assets or liabilities, vehicle finance, microfinance, or gold loans, occurs on a single platform. This will ensure a consistent experience across all services. Customers and users will benefit from this uniformity and consistency, leading to greater satisfaction.

How do you compare Equitas to your competitors in terms of technology adoption and execution?

I would say, in terms of adoption, we are definitely better. Regarding execution, as I mentioned, I still want to achieve much better speed. I am sure at least one competitor is currently faster than us. If I can improve our speed, then our speed and execution will catch up.