Transcript
Welcome back to Nutanix Cloud Interlock Podcast Season 2. Today I have the great privilege of being joined by my good friend, Mr. Maxim. Hello again. Hello. With us today in the studio, who leads the Enterprise Risk Management Technology for Standard Chartered Bank, Ankita Gupta. Welcome to the session. Hello. Thank you for having me here. It was great to meet you at our Hybrid Cloud Day the other day. Please introduce yourself to our audience. Tell us a little bit about what you do at SCB. Hi, I'm Ankita Gupta. I head the Enterprise Risk Management Technology in the bank. And with my diverse experience in Standard Chartered, I led the digital transformation and the AI enablement for the risk fraternity in the bank. Of course, with the evolving technology, there is a lot that I have learned through the years, which currently we are exploring from cloud migration to the AI enablement. And we are going to talk a lot more in this session. Thank you so much for joining us. You run a part of the business that I think, especially in a financial services organization that has probably two components to the risk. The first one will be like from a regulatory standpoint. And the second one that I often talk about is a branding and awareness, especially when we're seeing the rise of, say, the smaller digital banks, and then you've got the larger banks. What's your sort of blend in terms of how you see that risk profile from a regulatory versus brand? Do you take both into account? Which one, obviously, regulatory would come first? Yeah, no, it definitely has to be a balanced approach. You know, banking data is one of the most regulated data in the world, right? And we have to comply with GDPR and other regulations. For sure, branding is a major part of it. But as you correctly mentioned, regulation comes definitely first. And that is where with the digital banks as well, for Standard Chartered, we do have Mox Bank and Trust Bank. So that branding forms an important part as well. And when you think about that now, from a risk perspective, and with explosion of AI and different areas to adopt with AI, how do you see that as opposed to, say, cloud when the whole cloud first or cloud native push started maybe four or five years ago? Definitely cloud migration started much before the boom of AI. It was not that people were not using AI. There were organizations that were using AI, but the boom happened recently. Now, what we are seeing that there is a close connection between cloud migration and enabling ourselves for AI. First of all, I want to make it very clear that cloud isn't just an IT play. It's a business agility engine. And for financial organizations, it would mean better loans, higher investments, and happy customers, right? Now, a lot of AI evolutions and innovations are happening on cloud. Now, this is well understood that if we have to tap on those innovations and evolutions, we have to find a way to migrate to cloud. But to what extent? Keeping in mind the regulations, that's what a balance today financial organizations, I would say, are considering. And a lot of organizations have gone ahead in this as well. The well-known banks in the world, if you see, they are investing heavily on AI, on cloud. And that's where the financial organizations are even considering a lot of conscious cloud models, whether they should go on public cloud or private cloud or a hybrid cloud, public and private. Yeah. And how do you deal with the complexity of the bank of your scale? You're present in more than 50 markets, if I'm not wrong. Public cloud would not be present in all of these regions. So we can't move the data cross-board in most of the cases. How do you deal with that complexity? How do you build a model that allows you to manage that risk? We operate in 52 markets. And yes, it's a diverse organization, right? And definitely private cloud doesn't serve all the purpose. It is not a cookie cutter for us. So we definitely have a hybrid cloud approach, wherein we do have private cloud for a few of the countries where there are data residency challenges, especially around like India. We have some RBI mandates. So we do have private clouds where our core banking is going now. And then there are a lot of operations like the countries like Singapore, Hong Kong, which are more welcoming to going on the private cloud. There we are going on private cloud. But of course the caveat is with the right guardrails and the encryption of the PII and CII attributes. Yes, the move is happening. But back to your question, yes, private cloud doesn't serve everything. And from a point of... Often we talk here about how a single operating model, a single governance and security model is how you're going to be able to adopt it fast, make sure that what you are consuming in cloud is relevant and meets all the requirements of regulation. How has that journey changed for say someone as large as ASAB, where what we've seen in the past is, here's the on-prem team, they run the private cloud. And now we want to use a public cloud. Has that evolved? And have you merged those teams? The problem is not how fast we move to cloud. The problem is when we don't know how and why and where we are moving. So back to your question. Yes, so there is a very clear strategy that we are following on where to go on private and where to go on public. Now to tap on the skillset internally, we have merged these teams on a central team because that's where it's like a COE where the knowledge lies and they will be able to have an approach. If they have implemented for a previous project, they'll be able to further kind of fast track any future projects coming in. How this way of AI is different from the previous one. We used ML and fraud detection for ages in banking. Now we're slowly moving to the customer facing functions, to the engineering. Do you see different traction? Do you see different challenges or it's been sorted out? The biggest thing is how do we do all this? Making sure that the trust with customers is kind of still retained, right? Yes, for the fraud detection as well. We are using, we have seen in the industry using the AI which can detect some fraudulent activities and hence which I think a normal human being cannot detect. So for sure we have seen that fraud is an area which is a favorite for AI and which I would say AI has been implemented for it in the banking organizations much before we move to the right and move to the customer journey, right? Now, as I said that we are moving to the customer side as well, like there are advisory services, right? That are being rolled out on the various cloud providers. So this is an area where the organization, banking organizations are definitely moving towards. But then again, how much can we trust that advisory? What should be the human in loop? Or should we just have an agentic AI framework and leave it to the AI to decide? I think that's a part which banks are still kind of understanding where to go. Interesting point, especially around how you adopt that and who's going to adopt it. And I think understanding, and I was meeting another bank last week and they were talking about their current utilization of Bedrock on AI. And they said, what would that look like if we brought it back? And we spoke about the technology part, which is, it's not that difficult to be honest. However, it then quickly turned into, I don't have a team of data scientists. I don't have all these mountain of engineers that I can just put behind it that say the cloud providers do. Is that something, maybe not so much from a risk perspective, but when you're looking at implementing and when you're looking at changing the models, how are SCB adopting that? Are they saying, we're going to grow our team? This is a core principle of where we're going as an organization. So we're going to invest in growing our data scientists and AI capability, or are you leveraging external partners to provide that knowledge? So I think it's a hybrid and that's what is in the industry as well. See, I am very much into digital transformation. I'm a strong believer that we should not reinvent the wheels. If the world has developed it for you, they have a better capability. Let's go out in the market and accelerate your journey, right? Rather than trying to build everything in house. Whereas we also have to be very considerate and cautious that we need to build that AI and the technology muscle in the bank or in any organization. It's interesting you say that, because I think the first fear I saw as soon as AI, probably maybe the last two versions of OpenAI was, AI is just going to replace every staff member in our organization in five years. You spoke about productivity and financial gains on an area of data through AI that would have had probably zero impact on a workforce perspective. How do you see it? Whenever we talk about AI, this question definitely comes. It's the first one. But I always say that if you look at 1970s, when internet came up, 70% of the jobs, the skilled, the job families that exist today were not there then, right? So our whole job descriptions, our requirements in the organizations have evolved. And hence with AI coming in, I don't see that there will be a need of less people, but there will be an evolution of the demand of skills in the market. Secondly, I am, again, a firm believer, and that's what we have seen that with AI, you do more in same. That means a single person can do more stuff. I won't say you reduce the people, right? But you do more with the same person. Like for example, with our option of GitHub Copilot, we have seen for an engineer, there's a 30% efficiency and a productivity that comes up. I often talk about how getting hybrid cloud done right. It's not going to change your business, but it's going to provide you with the foundations for what is going to be the differentiator. And it sounds like you've really embraced that as well. And that whole digital transformation is going to be the pinnacle for differentiating, disrupting, opening into emerging markets. It's a really refreshing take, as opposed to I'm going to draw some pictures and write an email with Copilot or something like that. Yeah. Yeah. No, for sure. Cloud should be seen as a business agility engine, right? And that is where there are a lot of tools today, which we see are not available on-prem, right? And that is where if we have to tap on those tools, which are bringing in much more efficiency and productivity, we have to embrace cloud, right? And that is where slowly and steadily with the right guardrails and the right governance and risk management processes, we have to move ahead. And very frankly, over the years, we have seen that the regulators also are very much welcoming to the idea of cloud with the right guardrails. Are you finding that change with the regulatory bodies differently around Asia and APJ? For me, I've seen it be no cloud, no AI. Now it's, okay, you can have cloud, but how are you going to protect if that cloud fails? So I'm seeing them move. It's definitely not at the pace that is evolving with the technology, but it's getting better. Yeah. Yeah. There are regulators and especially for banking organizations, right? Where I said that no outage systems where a core banking runs is of prime importance. Otherwise it may have heavy fine for the banking, right? And that's where the regulators are too much, are very kind of cautious about that. How a banking organization is building a good disaster recovery muscle by having an inter-cloud flip, right? Because of late we have seen even the cloud providers are having outages. It is disrupting the businesses. So how do we make sure that if suppose an event happens in Azure, the core banking application is able to flip to AWS, right? And that's a tricky part. I would say that agility across clouds is not easy to achieve. It's not easy to achieve. And that's where I think banking organizations would need a lot of conscious effort to build in that. We have definitely able to achieve the geo-resiliency, right? But when it comes to inter-cloud resiliency, that's something we have to work on. And how that relates to the banks that born in a cloud, like the digital banks that has nothing, they don't have modern breaks, they don't have data centers. They just have a cloud. How they have to build this from the scratch? No, so agreed. Digital, as I said, with the right guardrails in place and with the right encryption in place, right? When you see the digital banks, I do see there are countries, mostly the digital banks are more prominent in the countries which are more open to cloud, right? But yes, with the digital banks, and this is my experience, that they have the right guardrails in place. Now, when it comes to more of an open banking, like the non-digital bank, where it operates in diverse market, that's where having that single technology that will, or the single setup, that will help all the locations and markets. That's where the challenge comes to scale up, right? Because the digital bank's footprint is much lesser. And I suppose the digital banks don't have this brick and mortar history of 20, 50, 100 years of applications and customers that they're building on. Do you think they're, I won't say the word lacks, but do you think they're a bit easier on the restrictions to try and promote the changes in the economy? Or is it really just, they're not sure? When it comes to, again, the transaction-related processes, I don't think the digital banks are even compromising there, right? Or even regulators are too easy on them, right? But as you correctly mentioned, they don't have a history of 100 years, 135 years. That's where they don't have legacy systems in place, right? So they don't have a legacy history to move to cloud, right? They know from this point of time, where they're already very much digitized, they have to move forward. Whereas if you see larger organizations with the history, they have a big legacy in place, running on mainframes and all that stuff, right? And there it becomes, as I said, to scale, very difficult that, okay, do I start from this day to go on cloud? Or do I need to take legacy? Usually for core banking and all, you just can't leave legacy on prem. Because the transactions are all historical for a client. You have to move with the whole data and the processes on cloud. I think that's where going from legacy, having that legacy flavor to move to cloud at scale becomes a challenge. Yeah, I understand. And Kira, I'm going to ask you one more final one, which is for organizations that don't have the staff, the resources, someone like yourself to help them on that digital transformation journey, how would you say they get started? What would be your number one piece of advice that you could give them when it comes to digital transformation and getting going? I would say start at least. Just start. Just start. Start with something which is not a very critical operation for your organization, right? At least you will build the nuances and that muscle and a know-how that how is it all going to span out if you have to scale, right? So I would say plan for the small, but at least start, right? I'm not saying that don't take the whole ecosystem in consideration that what is going to be your target state, but in your overall target state, try to prepare a plan which will show you smaller wins rather than having a two-year journey and just realizing at 1.5 years, oh, we did it, yeah, we are not doing it, right? So I would say engage with the right partners, the right skillset, and start off at least. Awesome. Awesome advice for everybody. I think Maxim and I say the same thing. Don't sit there and wait. Make a start because if you don't, someone else in your industry is definitely doing it. We know that. Yeah. Ankita, thank you so much for joining us today. Absolute pleasure to have you here and thank you for sharing your knowledge with us. Thank you so much. Thanks. Thanks for coming. Hopefully you found today's episode nice and insightful. Should you have any additional questions for the team, send them on through and we'll cover them off on the next episode of Nutanix Cloud Interlock, Ask Us Anything. This is the Nutanix Cloud Interlock podcast.
