AWS Financial Services Symposium 2025 | Rajesh Santhanam & Prabhakaran Pitchandi, Tata Consultancy Services

Clips
News
More from AWS Financial Services Symposium 2025

Rajesh Santhanam

Global Head AI, Cloud & Ecosystem, Financial Services

Tata Consultancy Services

Prabhakaran Pitchandi

Vice President and Global Head, Analytics & Insights, Financial Services

Tata Consultancy Services

Inside the AWS Financial Services’ AI playbook: 8 expert conversations from labs to ledgers

AI for financial services is no longer a distant vision — it’s reshaping how the industry thinks about risk, regulation, customer experience and innovation itself. From streamlining fraud detection to modernizing data architectures, financial institutions are leveraging AI not just to keep up, but to lead.As enterprise-grade tools mature and cloud platforms such as Amazon Web Services Inc. scale up their offerings, these organizations are deploying AI in tangible, transformative ways — and watching new value vectors emerge in real time.That momentum was on full display during the AWS Financial Services Symposium, where theCUBE captured a series of candid, expert-driven conversations

play_circle_outline The Evolution of AI in Financial Services: From Experimentation to Value Creation

play_circle_outline Blockers for AI adoption: risks, trust barriers

play_circle_outline Value creation through business process redesign and customer-centric approaches

play_circle_outline AI in IT: modernization, code conversions, reducing technical debt

play_circle_outline Balancing Risk Reduction and Profit in Financial Services: The Role of Data and AI

Info
Transcript

Rajesh Santhanam & Prabhakaran Pitchandi, Tata Consultancy Services

Rajesh Santhanam

Global Head AI, Cloud & Ecosystem, Financial Services Tata Consultancy Services

Prabhakaran Pitchandi

Vice President and Global Head, Analytics & Insights, Financial Services Tata Consultancy Services

In this AWS Financial Services Symposium segment, theCUBE’s Dave Vellante sits down with TCS leaders Rajesh Santhanam, global head of AI, cloud and ecosystem, Financial Services, at Tata Consultancy Services, and Prab Pitchandi, VP and global head of analytics and insights, Financial Services, at Tata Consultancy Services, to explore how AI – and especially gen AI -– is reshaping the financial landscape. The conversation dives into the modernization of legacy systems, regulatory realities and real-world use cases driving scalable, secure innovation.

Sa... Read more

explore Keep Exploring

What has been the evolution of AI usage within the financial services industry over the past few years? add

What are some of the challenges and concerns faced by regulated entities, such as banks, when considering the adoption of AI technology? add

What are some key ways in which AI can create business value in various industries? add

What impact is AI having on IT modernization and code conversions for clients? add

What are the four different parts of the framework for preparing data for AI usage in financial services organizations? add

bolt Powered by CUBE AI

Rajesh Santhanam & Prabhakaran Pitchandi, Tata Consultancy Services

search

Dave Vellante

>> Hi, everybody. Welcome back to New York City. My name is Dave Vellante and you're watching theCUBE's day-Long coverage of AWS Financial Services Symposium. As you know, financial services oftentimes leads the adoption of technology. It wasn't necessarily the case with cloud, but it most certainly is with AI and other innovations like agentic. Super excited to have Tata Consulting Services back on theCUBE, Rajesh and Prabh. Rajesh is focused on AI and cloud and ecosystems, specifically in financial services, and Prabh, he's a data guy with analytics and insights around financial services. Well, gentlemen, welcome to theCUBE. It's good to have you.

Prabhakaran Pitchandi

>> Thank you. Good to be here.

Dave Vellante

>> So you heard my comment up front. I mean, initially I remember the days when the financial services companies were saying, "We're not doing anything in the cloud," and it turned out they're very aggressive in the cloud and they realized they can change the operational model. So my question is, related to AI adoption, what is the current state of AI adoption in financial services? Who wants to start?

Prabhakaran Pitchandi

>> Well, I can go first. When you look at a couple of years back, where we all started within the financial services, was fear of being missed out, and so we started with experimentation and lots of experimentation was done at that point in time. Then, now move on to the next phase in terms of how we can create business value and how we can create efficiencies out of it. The interesting thing was, because AI is so democratized in the world as it is, the business looks at it and say, "Oh, I can do it myself and then go and ask a question on ChatGPT and get an answer, so why not we start leveraging AI for our business?" That's the kind of push that has come from the business in terms of using AI for value creation. So where it stands today, if you look at it, I will put it in different ways. There are some firms who are still in experimentation stages and some have actually moved on to the production and started creating value. And the leaders who actually have done and gone beyond experimentation, the way they're looking at was, first, do I have a committee in place to manage controls and governance? Do I have a platform? How do I go about creating platforms? And then how do I onboard use cases and how do I create value? So that's the kind of evolution that the banks and insurance have been going up with. If I just have to add a little bit more in terms of where the adoption level is high, when you look at AI and ML, that is not new. For a bank, doing statistical models is not new, because banks are models-driven, but the AI/ML has been more in terms of risk and AML and some in the sales and marketing, looking at some personalized insights, now that is going beyond and covering a whole host of things. How do you look at summarization of information, extraction of information in lending processes or customer agents in the customer experience processes? And now moving beyond that into more specifically on creating value chain transformation. So that's the kind of journey that our customers are in this space.

Rajesh Santhanam

>> And my take on that is, AI has been used, AI/ML, traditional AI has been used in financial service for a number of years, so it's not new to the clients in the industry. However, when GenAI burst onto the scene, that's when the interest really spiked in terms of AI and AI adoption. So from the last two years then there has been a lot of experimentation and pilots that have been done by various clients. We're now seeing AI becoming more and more mainstream. So it's being used in areas like customer support, a lot of internal use cases to help employees find information, and it's also getting more to the front end where clients are also getting exposed to AI and GenAI in some ways. So it's becoming more and more mainstream from the last two years where it's been experimentation and pilots. So that's-

Dave Vellante

>> Using an AWS term, what are you seeing as the blockers for AI adoption? I mean, you've still got a lot of systems that are not in the cloud and maybe never will be in the cloud. I know we talked to a lot of banks who say, "Well, we've got to bring... We're going to build our own AI stack out and we don't really have the skills, we don't really have the stack, the cloud actually has all the tooling." So that latency is another thing. Sovereign AI, with all this geopolitical stuff going on, what do you see as the big blockers to-

Rajesh Santhanam

>> Yeah, some of the blockers have been the risks associated with any new technology like this, because when people jump into this, they don't know what the risks are and it's not fully understood. And over the last two, two and a half years, the risks have been better understood by our clients. So there is more awareness and a better systematic approach towards addressing those risks. So we have something called a responsible AI framework that we encourage our clients to adopt, which takes care of the privacy, security, and other aspects that need to be handled before a regulated industry can adopt this at scale.

Prabhakaran Pitchandi

>> I would add, one critical challenge that the regulated entities like our customers in the banks face is a trust barrier. And that is a big thing. How much can I trust AI to ensure the privacy is maintained? How much I can trust AI in ensuring there's no hallucinations? How much I can trust that the answers, they're transparent and there's a lineage and traceability back to it? So this, in spite of there are so much of solutions and improvement in AWS themselves as a platform on the responsibility and we have a framework for that, as Rajesh mentioned, still there is a leap of faith that some of the customers have to make there. And some of the things that they're currently doing to make the leap of faith, to ensure that there is proper governance committees there, the risk committees are in place, there is proper guardrails and controls are in place, properly understand what the ethical use of data and how the bias is managed. So all those things are coming into play to ensure that we can overcome the trust barrier.

Dave Vellante

>> TCS is a world-class consultancy. It's actually a remarkable story, the evolution of your company, and you're all about delivering value, helping customers achieve value, and that's what it's all about at the end of the day. I'm hearing that a lot of the value today in terms of the blockers we just talked about is risk reduction. But of course, banks and financial services are all about money. They want to make money too. So how has the value equation, how is it shifting and ebbing and what's your approach to helping customers realize that value?

Prabhakaran Pitchandi

>> So as you rightly said, today many of the use cases are in productivity and inefficiencies from a GenAI perspective. From an AI perspective, there has been lots of use cases that creates value. For example, personalized insights, helping in sales and marketing, providing real-time pricing and more optimized pricing that can link to the P&L. So from an AI perspective, quite a lot of business value creation is there. GenAI is just in the beginning, and so it's more productivity and efficiency. The way we approach our clients is, it is good to start with use case-specific approach, because you try and you try different things and how it works, but you have to leapfrog into value chain transformation. So when you look at the underwriting value chain or we look at the end lending value chain or claims value chain, when you start looking at transforming the entire value chain, then you create more business value. That is number one. Number two is, we try to move more closer to the customers and move closer to the front office. For example, in the capital markets world, you can use AI and GenAI to analyze a whole host of data, the data that is there in structured pharma and unstructured form to create insights that help in your investment decisions that are helping your trading decisions. And so that movement is happening in terms of starting to look at using these technologies in the front office side to create value as well. So in a nutshell, I would say value chain transformation is very important, and always look at any use case or any change that you're transforming, what's the business value, and move more towards customer, the front office.

Dave Vellante

>> I remember customers used to say that just lifting and shifting to the cloud, it's nice, get rid of some IT, but it doesn't change the operating model. And now you're seeing major changes in the operating model. What you just described, Prabh, is pretty transformative. Is this business process redesigned? 2.0 or 3.0? But you had a comment. No, please, go ahead.

Rajesh Santhanam

>> Yeah, so our clients have been looking at the return on investment on these AI engagements that launched inside the institute, so it has reduced dramatically in terms of the costs of inference and the overall AI costs are plunging by the day. The models are getting smaller and more efficient and we are also seeing a number of our clients experiment with small language models as opposed to the big ones. So that's more tuned towards certain tasks. It's able to reduce hallucinations, it's able to deliver results at a much lower cost. So we are looking at ways like this to improve the value of AI to our clients, and the business cases are stacking up better because of these factors, like plunging costs of AI as well as utilization of models appropriate for the purpose.

Dave Vellante

>> Yeah, we've definitely seen the adoption of small language models. I mean, we saw the whole DeepSeek thing just sent the whole market going crazy, but it was a sort of seminal moment in the industry and really underscored that. And we seem to be going beyond sort of RAG-based chatbots. It's like that was such a hot topic and then all of a sudden it's like, no, it's all about agentic. So where do you see this industry, particularly financial services and adoption of AI, GenAI, agentic in the next, let's say two to three years?

Prabhakaran Pitchandi

>> Well, AI is moving towards better maturity. And so, as you said, it started with more of a chatbot and you have a context, you have a prompt. Then okay, that information is not enough, because you give a prompt to a general purpose LLM, it gives an answer, not enough. You have to give more context to it, fine, RAG model. And then comes, okay, but then it does not have a state, it does not have a memory, it does not know how to call a function to make it take an action. Then comes the agentic AI with all the frills in terms of making AI as an application in itself. That is the feature. And why I feel that is a feature, is when you start looking at transforming value chains, it is not just one task, it is a combination of tasks. And so, when you start looking at everything as multiple agents and orchestrated in a way, either you can orchestrate in a probabilistic way or deterministic way, in other words an autonomous way or in a much more SOP-based way, but in respect to that, you have multiple agents working together with a particular goal or a business objective in mind, then you start to create more value for the business. That is the direction that we are traveling towards.

Rajesh Santhanam

>> So yeah, agentic AI is a big step up from what we've seen with the traditional foundation models and the LLMs. So the capabilities that they bring, the deep research abilities that they're able to offer to, say analysts and researchers in financial services industry has proved to be of immense value. So agentic AI, we see is the future, at least in the foreseeable months to come. Then we might have some other big change which comes about, but at the moment in 2025, we've seen so far that agentic AI and multi-agent frameworks are the talk of town. A number of our clients have been implementing these. And with Amazon and AWS, we've been using bedrock agents as one of the ways to bring agentic AI to life.

Dave Vellante

>> What are you seeing in terms of this core IT? I mean, as I said before, a lot of these systems have not moved into the cloud. The buzzword is modernization, but a lot of the modernization has occurred in the cloud. And maybe you buy a new server, because there's a business case to do so, whether it's energy or the depreciation cycle has gone through, but there's again still a lot of stuff on-prem. How do you see AI in IT? Are we going to move those workloads into the cloud? Many people tell us, "No, we're probably not. We're going to maybe bring AI to the data, maybe build our own AI stacks, maybe there's an outpost." You certainly see Amazon putting more emphasis on that, you see other sort of on-prem vendors. But all the action is in the cloud today, in AI, I mean it seems that way, but what do you think is happening? What do you see happening within AI in IT?

Rajesh Santhanam

>> Yeah, so AI and IT is proving to be one of the better use cases that we've seen for GenAI, because GenAI has got an inbuilt feature, which is also a bug, which is hallucinations. Now, in the whole software development lifecycle, we have an inherent mechanism to correct that hallucination and the risks associated with that. So it's a fantastic use case for AI use and technology in IT. So that's where we see a lot of traction. What AI is also doing now is helping our clients produce the technical debt they've accumulated over the years. It's helping with modernization programs, which were so far not possible because the costs were just too high and it took too long. Now, AI is making a huge difference to those areas. So it's having a huge impact in IT in terms of modernization, code conversions. For instance, we've used Amazon Q Developer with one of our clients to rapidly migrate them off an old version of Java onto new versions of Java. Similarly, across multiple technologies, we are able to move our clients forward to modern technologies at pace, at a lower risk and a much, much lower-

Dave Vellante

>> As an example of driving developer productivity, we had Marc Benioff on recently and he was saying that their engineering teams are getting 50% improvement. Probably that's up from 30% late last year. It's just incredible. And his point was it's going to continue. There's a data element here too. I mean, we really haven't touched deeply on your world, which is data. I mean, without quality data, you're not going to really succeed to having governed agents. So what is the state of data in financial services? Obviously a lot of structured data, but now we've got all this unstructured data coming in. Financial services organizations, in your view, are they prepared for this agentic era because they are so data-centric? Or is there still a lot of work to be done?

Prabhakaran Pitchandi

>> I would put the organizations in two buckets. The likes of market infrastructure firms and data providers where their business is data, they are making data sets available, AI-ready type of data sets available to be consumed by the consumers, like the banks and insurance firms. So that is one. And then comes the banks and insurance themselves. Now, when you look at data for AI and how to make it ready, I would put it in a framework and there are four different parts to that framework. Number one is use case-specific. Now, you need to have data sets that are readily available to a particular use case, so that the various scenarios are covered, various patterns are covered, various seasonalities are covered. So those are various elements that you need to cover to ensure that the right data sets are available for AI. That's the one. Second, the whole host of data management, and this is not new. We have been trying to solve the data problem for a while, and now with AI it has just come to the fore. Our customers have been investing for a couple of years, so well advanced in the stage, but some more way to go. What I mean by that is, having the right governance in place, having the stewardship in place, having metadata, having lineage, having the business glossary, the entire nine yards from the data management perspective. And that is very important for data for AI, and also, AI can help in achieving that faster as well. So that's the second element. The third element to it from when it comes to data for AI, is all this privacy and all the security in place, make sure the ethics and bias and everything is taken care of. So that is another important element. And the final point is to make sure that it is repeatable so that you have proper governance and controls. So it's not one-off, it is a continuous process. Today, there is a scenario, tomorrow the scenario changes, you're packing arrays, how are you going to reach in the model? You should have data to be able to manage that. So these are the four different elements of the framework that we normally advise our clients to ensure that we move and get data ready for AI. Easily said than done, because it just requires a lot of investment, a lot of heavy lifting and hard work to be able to achieve this.

Dave Vellante

>> And it really is a prerequisite. Last question, Rajesh, I'll start with you. What's the number one piece of advice you'd give folks, executives in the financial services industry that they should be focused on? It could be things to avoid as well.

Rajesh Santhanam

>> Yeah, I would say take an iterative approach, build on successes that you've had. Because what we've also seen is clients go too fast too quickly and have to reverse some of the decisions they've taken with regard to AI adoption. So I would say take a measured approach, utilize the full capabilities of what the technology can offer us today. And then as it matures, incrementally exploit the new capabilities of-

Dave Vellante

>> Avoid that cul-de-sac. You don't want to go down the dead end road and have to come back. Prabh, what would you say?

Prabhakaran Pitchandi

>> I'd say be bold and let your creative side really... Be on the top of your creative side and innovation. At the same time, ensure that you are able to bring the entire organization into play. You have the right platform, right controls, right scaling in place so the innovation can be monetized. I would like to add just one more point as well. To do this, we have platforms, we have a platform WisdomNext, which actually helps in accelerating this to help from innovation to creating value quickly as well. So I would also advise to ensure that you work with partners, because they have a broader exposure and will bring in the right expertise to help you move faster.

Dave Vellante

>> And it's a proven framework that you've done. It's not just something you've made up, you've applied it at many, many organizations. Guys, thanks so much for coming on theCUBE. Really appreciate it.

Prabhakaran Pitchandi

>> It's a pleasure. Thank you.

Dave Vellante

>> Great perspectives and thank you for watching. This is Dave Vellante at the AWS Financial Services Symposium in New York City. You're watching theCUBE. We'll be right back right after this short break.