Aman Sardana and Vijay Kumar Soni on Transforming Financial Systems with Hybrid Cloud | swampUP 2025

Hi, everyone. Welcome back here to our day two coverage of Jfr Swamp Up event in a beautiful Napa Valley. We're at the Meritage Resort, and if you've never been here, highly, highly recommend.

It is a full on resort and spa with wineries on and vineyards on premises. It's really a lot of fun. I I really do, do recommend it.

Talia. Um, let me introduce you to our next guest, though. We have two guests here.

We have Amman, sna, and I hopefully get this right. Vijay Kumar. Hey.

Hi, Ahman and Vijay work for a large financial, financial institution. And they're here talking about, you know, in a hybrid AI world, how, you know, how large financial institutions can get a hold of their, or it can, you know, manage their CICD pipeline, their software, supply chain, security of it, and everything else. So Armand is closest to me here, Vij to my far left.

Gentlemen, welcome to Text Strong tv. It's great to have you on here. Yeah, Thank you so much.

Yeah, we are, we appreciate the opportunity to be here and in this beautiful Napa Valley. Yeah, this beautiful day. There's worst, worst places to be.

Sure. Um, so let, let us talk a little bit about, you guys presented, actually. Yes.

Yeah. Yes, we did. So if you wouldn't mind share with the audience a little bit about your presentation.

Sure. So, uh, we gave a talk yesterday. It was, uh, a good opportunity to come here and, and present in front of like a large crowd talking about like how, uh, a large financial institution might do, uh, uh, a resident release of their workload, especially in a hybrid cloud setup where the complexities are very different, right?

I mean, uh, uh, typically like companies, they start small, uh, going OnPrem, especially in large financial institution, uh, there's a lot of legacy stuff that still runs on-prem, right? I mean, uh, especially in bank financial domain. Uh, companies started with building the infrastructure, uh, in eighties, nineties, early nineties.

And that infrastructure still runs on-prem. But the challenge is that, uh, in order to compete in the marketplace, uh, companies have to evolve and, and make sure that, uh, they, they are able to seamlessly integrate with FinTech, um, especially like the, the new age companies, modern age companies. And there is a pull, like on, uh, on the other side to build modern software.

And that brings a unique opportunity as well for the companies to start looking at how they can modernize their existing workload. But the challenge is it's very hard for the companies to migrate everything and in one single shot. So the strategic choice that company has is to invest in hybrid cloud capabilities, which doesn't come for free.

I mean, there are challenges, uh, that we can talk about, but definitely, uh, there are a lot of opportunities that show up with hybrid cloud where companies can scale, they have a capacity burst. Um, and then a lot of like where advantages when it comes to like innovation, making use of like, um, a lot of like different, like new technologies that are being developed and, and are cloud first. So that kind of like bridges sort of like the gap with hybrid cloud, where companies can still run the critical workload on-prem, but then at the same time they've find a way to sort of like, uh, gradually move, uh, to cloud in a phased approach in a phased manner.

Excellent. You know, I, I think, well, I'll speak for myself. I'm not gonna speak for everyone.

That's kind of the way I always envisioned it, right? That you don't, I remember, I, I was at an IBM conference eight years ago, seven, eight years ago, talking to the CIO of Hertz, you know, the car rental company, and, and he said they decided to go to the cloud. And what they did is they built from, from scratch all new cloud-based applications, and soon as they had those up and running, they literally just shut down their old on-prem data center applications and switched over.

Well, it was a disaster. Kurtz wound up filing bankruptcy, if you remember a few years back. Yeah.

He lost his job. Yeah. And, and so I always knew that that was not the way to do a cloud migration, right?

You don't do it that way. You had to do it over time, and you're gonna have this hybrid model and that, and there's nothing that matter. That hybrid model may last forever, quite frankly.

There's nothing saying to get, move everything into the public cloud, right? Yeah. I mean, if you look at, right, we are kind of home of innovation right here in, uh, California, and as you go from Napa Valley to south, you start seeing the innovation every, every mile, every minute, right?

Yeah. So if you look at this space and the, not just fintechs, the large electronics company, you know, the apples and Googles, how they innovated in last, uh, you know, one and a half, um, uh, decade. And if you look at all these things happening together, they're integrating with financial institution, they're integrating with healthcare and whatnot, right?

So if you see the innovation is happening and financial institution cannot be left behind, right? So they need to be, they need to make sure that the customer experience is retained. Customer wants everything on the fingertips.

So all this digital payments coming together, it's, it's, it cannot it stay on the legacy because legacy may not meet the customer expectations. So I, bridging the gap requires, you know, the resiliency everywhere, right? And we have a kind of, uh, environment where customer is at the center and everything needs to be delivered quickly to the customer.

So that's where the, the cloud and the DevOps and automation comes into the play and the finances services are catching up if they were behind, but they are not too far from getting, you know, no, uh, to the space where they're supposed to. So I, I'll tell you a a funny, not funny, but I'll tell you an observation I've had, right? com, it was 2013, so this is 12 years already.

Yeah. And, um, yes, traditionally, Finn finops financial institutions being conservative and highly regulated, don't necessarily have bleeding edge, right? They make sure something's tried and proven before they, they leap into it.

But when it came to DevOps and then DevSecOps, right? I saw financial institutions take the lead. And the reason for it was, quite frankly, where they were, was not a good space from a security point.

They were being attacked and hacked, and they had to do something. They had to get better control of that, number one. Number two, the market was changing.

You had this new breed of financial institution that was, was born in the cloud, that was born on the internet that was more nimble, right? Because customers, you know, when I was a little boy, we didn't have ATMs even yet. You had to write a check.

You wait online at the bank to the teller. Yeah. Yeah.

And they would give you the money. ATMs came out, but then apps came out on the phone, the cell phone, right? Yeah.

It just changed it. All of a sudden people were, how often do you actually go in the bank anymore? How often do you deal with the human at your financial institution group?

Only if there's a problem, basically. Yeah. So as a result of this, I, I think the market flipped finance became a leader in customer, you know, giving the customer what they want, when they want, how they want, which was primarily through an app.

Yeah. Not through a brick and mortar. Mm-hmm.

And, um, so my experiences is that they have led the way in DevOps, they have led the way in a lot of these innovations. However, they still gotta worry about highly reg being in a highly regulated industry. Right.

Let me though, ask you again, you, you made this move into this hybrid environment. Do you envision going total cloud or will you always, you know, I look a a lot of financial institutions still also have mainframes, right? Yeah.

And you're not gonna move, you know, to walk away from your mainframe. Yeah. That's a statement, right?

Right. That, that that's, that's not easy. Yeah.

com and in text drawing, they have, you know, they maintain a sy a system of record and a system of engagement, right? Yeah. Record being usually a mainframe at a data center somewhere, and engagement may be up in the cloud, right?

Is that how you guys envision your go forward? Yeah, I mean, the way I say is that, never say never. Okay.

But at the same time, as an organization, and depending upon how you are set up, you build the strategy and the strategy as the time passes because of the other factors. You know, you may plan and say, okay, I want to move into cloud, you know, but as you see the opportunity, you could probably shift and then get rid of the on-prem, or you think that Yes, probably in the long term strategy, I have the midterm milestone to determine whether I can get rid of or I still need to maintain. So the ag organizations build their strategy depending on the size of organization, depending on the customers they are serving, depending upon the partners they do have, they make the strategy in general.

There are a lot of organizations they can get rid of this, um, on-prem setup just because the way they are set up and the customers they are serving in probably our case. And there are some similar cases, probably it might be, uh, time to see, you know, in next couple of years or few years to see that, oh, now we have moved enough, probably it's time to, you know, get completely, uh, other side of the bridge. So that, that's kind of strategy, you know, as, as, as we go.

So, so short answer is we are yet to see that how the strategy evolves to, you know, get to the fully cloud and we stay onto hybrid mode. Yeah. And just to add to like what we just said, I think, which was very well said, um, I think companies also have to make sure that they're doing it in a very sustainable manner, right?

So, um, I think we just alluded a little bit, uh, on that as well, that it should not be like a big bank sort of like move to cloud. Uh, of course, I mean, it's a learning opportunity for a larger institutions, right? I mean, where, uh, they might have like thousands and thousands of applications running on-prem.

Um, they have like few applications running in cloud. Now there are challenges that show up, right? I mean, you might have a data sitting on-prem and there might be regulations right around how you store the data.

Uh, you might have to be in compliance with, with your data, right? So, uh, sometimes you might want to move the compute to cloud, uh, to the cloud, but then you might have, uh, some restrictions that you cannot move the data to cloud. Uh, there might be some, uh, innovation that might have to be done in order to be able to move the data to cloud.

Uh, but I think that's probably, uh, uh, kind of like a, uh, uh, an approach that company can take where they can take some steps in a phased approach before they finally land in, in a, in a, in a state where they can declare that they're fully running into the cloud. But I think that that should be probably the mantra for any company. I mean, that could be the north go north star for the company, but again, there are various factors that can, that could influence that journey, uh, towards that North Star.

Agreed. Yeah, agreed. You know, another phenomenon, so I hybrid to me was always obvious.

Yeah. You know, what wasn't obvious to me, multi-cloud, right? So do you see an institution like yours going multi-cloud using more than one hyperscaler cloud provider?

Yeah. Maybe I can just quickly, uh, throw some light on that. So definitely there are like a lot of different cloud providers, public cloud provider, and every cloud provider has their own sort of like niche, right?

Yep. Uh, you can look at my is Microsoft Azure, uh, Google Cloud, Amazon Web Services, uh, all of these like big cloud providers, I think, uh, they, they all are innovating. Um, they are sort of equally good.

Uh, but some cloud providers have services that pro that pro that probably might be like more niche and more sort of like cater to a certain sort of like, uh, uh, industry, right? So, um, I think in general, companies are sort of like investing, uh, in one major cloud provider, but also being at the same time cautious about the fact that, uh, there has to be some sort of like a risk management approach that has to take Well, you don't want to locked In. Yeah, you don't want already locked in.

But at the same time, uh, it's also a conversation around like the cost management. Uh, you don't wanna put all your eggs in one basket. Yeah, that is true.

At the same time, if you divert too much, then uh, you are also kind of like, uh, not reaping the full benefit of like having a cost, sort of like optimization by putting more workload in, in one single cloud provider and then kind of like being able to negotiate, uh, a price point that serves you well, right? So, uh, it kind of like, it can go either way, but then of course, I mean, you have to have a balanced approach as to like how much you diversify, but at the same time how much you concentrate and then kinda like get the cost optimization. Yeah.

To add to, among what he said, like if you look at probably no organization, which is using a service as a SaaS is insulated from multiple cloud, you know, um, uh, environment, probably directly running its own workload or probably via a service provider, there could be likelihood that you are using underneath, right? But the important point is that if, depending upon the size of the company, the customer, it is serving the geographic location and pricing point, the organization chooses to be, you know, into the multi-cloud versus single cloud now, and then the on-prem, right? So if you look at the landscape, they are, you know, in the multi-cloud strategy probably is better in certain cases when you are like global company and then you are serving the, you know, different type of, uh, um, customers and services and then, you know, um, the kind of, uh, product you are offering.

And in that case, sometimes you may, you may have, we not have a choice other than going to multi-cloud. So those factors come in. But if you are like a small organization, probably you may opt in to have probably two providers, and probably you are happy with that.

And so you can negotiate the pricing and do other things. But at the same time, if you are, say if you're going to a CF aspect reason, and you do not have that cloud provider capability, you are bound to have used another cloud provider, which is already there because you want to serve your customer. So those factors kick in.

But certainly from the, from the financial institution or you know, the global companies which are serving the customer, they probably make those, uh, smart choices depending upon their business strategy, uh, in global landscape. I love it. Gentlemen.

I apologize. We got a lot of background noise. Hopefully it's coming out well in microphones.

Um, last subject, ai, how's that playing into this? I think no company is shielded from ai. Uh, I think that's, given that in order to stay ahead, uh, companies have to make investments into ai, but again, it at the same time, financial institution, uh, banking and all, like different industries have to be very cautious about like how they use AI in a responsible manner.

I think one of the biggest challenges, uh, especially when it, when, when, like, when it, when it relates to like financial matters, right? Um, AI can sometimes make a decision that may not, uh, be very well in kind of like aligned with regulators, right? So in general, um, AI has a lot of potential, but I think there is still, uh, some work that needs to be done to make sure that AI is used in a very, very responsible manner.

And there are use cases where, uh, AI can be easily used, right? I mean, uh, we have seen financial institution investing like in deploying AI in call center applications just to help agents serve better, uh, for the customer. But I think what we have seen, like in the industry in general, is AI being adopted in a core sort of like processing is still sort of like something that has to be, uh, uh, has to be, well, kind of like, um, uh, there has, I mean, it requires some sort of like investigation as to how it can be incorporated in a responsible manner.

Because I mean, anything that you do has to be explainable, right? And sometimes the challenge with AI is that it, it's not very conducive to explain like how decisions were made, like, uh, by the ai Yeah. For certain industry it might be simpler, you know, if you go to manufacturing, it might be simpler because you are doing mechanical in most of the cases.

So it's a simple, it's a repeatable process, you know, you're not making a lot of decisions there. But if you go to the healthcare or, you know, uh, into the spaces where humans are involved, you know, decisions are involved, finances are involved. I think it's, it's, it's very, you know, the area where it's not easy to Not in the highly regulated Can go move forward with That highly regulated.

Yeah, exactly. So, agreed, those considerations, you know, are you always there? But as AI matures, who knows, you know, in next few years things may change differently.

You know, what if the public, if your customers demand that sort of functionality, you're not gonna have a choice. That is true. I mean that, I mean, customer comes first.

That's a Customer led business. Yes. Yes.

Yep. Anyway. Hey guys, thank you for coming on you For having us today.

Thank you. Good. Swamp up.

Thank You. Yeah, thank you so much. Yeah.

All right. Vj, Oman. Aman, yes.

Here on Techstrong tv. We're gonna take a break. We'll be back in a moment.