Name: 09122025 Cognizant
Uploaded: 2025-12-09T08:33:40.267Z
Duration: 46 min 26 s
Description: 09122025 Cognizant

Transcript for "09122025 Cognizant": Good afternoon, everybody, and welcome to this exclusive session for banking leaders brought to you by Cognizant and Microsoft. AgenTek AI is moving fast, and banks are shifting from isolated pilots to full scale enterprise transformation. This is about more than just efficiency. It's about fundamentally reinventing business models to boost performance. Today, we tackle the critical question. How do we achieve scalable success? We'll define agentic banking, cover the essential foundations, explore the strategic pathways to scale, and finally, look at how agentic AI intersects with trends like DLT and tokenization. I'm Ella Wilkinson, broadcast editor for FinTech Magazine, and I'll be your host today. And I'm joined by an expert panel. It is my pleasure to introduce John DeGamaRose, a key executive from Cognizant, Patricia Mann from Microsoft, focusing on cloud and AI strategy, and Robert Benio, who drives large scale transformation initiatives with a keen eye on AI. So I've given you a small introduction there, but who better to introduce yourselves than you guys? So, Patrice, you're on my screen right now, so let's start with you. Yeah. Thank you. So this is, Patrice from Microsoft. I'm regional leader for our EMI and financial services business. And with my team of industry advisers across, our time zone, we're focusing on large industry transformation powered by cloud and AI with the aim to deliver business outcomes. Fantastic. Robert? Thank you. Good afternoon. I'm a BFS market maker and solution leader for what we would call global growth markets. I'm a GRC hobbyist, so governance, risk, and compliance. So for all the futurism, I'm always focused on the resiliency and risk management elements, and obviously, the views are shaped and informed by twenty five actually, twenty nine years in banking. Brilliant. And John, finally. Thank you, Ella. Good afternoon, everyone. So John DeGumerez, I lead our banking and financial services business across our international markets, which, covers EMEA and APJ. I've been in banking for over thirty years now and looking forward to the discussion. Brilliant. So let's actually just dive straight in, with the easiest question of all. What exactly is agentic banking? Patrice, if we start with you. Oh, you think that's the easy well, the question is easy, but then the let's wait for the for the answer and the But the answer can be complicated. Exactly. No. If I if I may try a definition, I would say Adjantic banking is probably human led and AI driven, leveraging autonomous AI agent that can perceive, decide, and not towards define goals across banking operation. So it's really about moving beyond traditional automation to proactive self directed workflows. There's a core difference with traditional tech and conventional AI, and this one is that we are shifting from assistance with task to the execution of multi step action independently, adapting to new or extended datasets and collaborating with other systems. I think in terms of purpose, it's really about enriching employee experience, it's about reinventing customer engagement, it's about reshaping business process and finally, bending the curve on innovation. There needs, of course, new capabilities for that. It's about autonomy, so be able to act without constant human problems, I would say. It's about adaptability, which means learning and adjust strategy dynamically. And it's about coordination, because the orchestration across multiple agent and system will be key. So as a summary, I would kind of say that agenting banking is reinventing banking models by combining efficiency, personalization, and compliance, positioning bank for digital first and outcome driven operation. This is to me the road to become what I would call a frontier firm. Very concise answer that sums up absolutely everything we're gonna talk about today. John, did you have anything to add? I thought I thought that was an excellent, response by, actually, Patrice and, agree with his, you know, summary. I think I think for me, it is that element where it allows you to have a multi agent component to the model where, you know, that that reasoning, the performing of exercises, the orchestration, I think, is key. And it really just focuses on a value stream rather than an old process and really gives you a hybrid model where you're you've got the human and the and the agents working, to together. But the agents are really doing the heavy lifting and basically providing efficiency into the model where where the humans are, you know, providing the oversight. So, you know, it is the model of the future, and you can see at the moment that, many of our clients are deploying agents now and, actually getting the benefits from them. So let's talk about that business model reinvention then. So how can AgenTic AI move banking beyond just process efficiency to genuine business model reinvention? Robert, if we throw to you. Yeah. Thank you for that help. You know, I guess one message would be, every member of the enter of the enterprise has gotta be a Jetix superpower. At the moment, I think a lot of organizations are still focused on what you all remember and recall from some of our blogs, maybe Vector one. They're all thinking about how do I make the technology development life cycle run faster or how do I automate a process. If you want to move beyond that process efficiency piece and go into genuine business model, Ria mentioned, we'll talk about that a little bit later and we've talked about it in the past. It's kind of the vector too. That's what we gave a there was a persona we created and we sort of said that the the person there, we said, was Sarah. And Sarah was the business owner, and she would use, you know, you have to imagine relationship managers in a bank, frontline customer service support members. They're all going to have a Gentic supporting them either in one way of either assisting, augmenting them. So assistance in our mind is, getting the support. The augmentation is is actually the answer may be even crafted or recrafted as it's being spoken by customer service agent. And then the rest may just be automation, which is there's an intervention done by, what is a synthetic agent coming in. It would be you know, it's agentic is, by default, not a tooling that will be driven by ops or the way that RPA or IPA was, but across the enterprise. And if you see how the AI leaders are instantiating AI in their organization, and by AI leaders, I sort of say they have the the AI organizations reporting to the CEO, not necessarily to the CIO and the CTO. All that coming together is to sort of say, it will only happen when the organization starts to prioritize it at the enterprise level and not at the IT level, not at the ops level. And John? Yeah. I mean, I I think I think with with Adjantic Banking, I I think it both it covers both technology and operations. It's not only an operations angle here, and it it feels to me that you've got to go back to looking at the value stream rather than going back to the existing process, and try and enhance the process, which is what the RPA did. So I feel that, in order to get a real benefit, you've almost got to reinvent the value stream across, you know, function or or industry vertical, you know, within the banks, to make sure that, you know, you've got a different view in terms of how to reason and how to process and how to execute in a multi orchestration layer with both, you know, the agent and the human. So so I think I think there's an element where it covers both tech and ops, and you'll you'll have a a a a concept, especially in tech, where you've got an ADLC component to think about. But don't go back and apply it to an existing process because you won't get the value. Yeah. GenTIC is coming as we need we need to obviously learn how to embrace that. And Robert, like he's currently said, we will cover that more later in this webinar. So let's define this further then. So AI is a broad topic. And just like intelligence has a definition and dimensions, what do you think matters most with this? John, you're still on my screen, so let's throw it to you. So I think, where where we have the evolution of of of AI and, you know, Gen AI, which is the fact of the productionization of intelligence, which is we assist me in the whole series of things. And I I think that component part is still very important. Where as we're moving to an agentic world where, you know, it still works with AI and Nigene AI, but it's the it's the element of now you have to perform something. Now you have to do something for me. And I and I think those components those three component parts of AI still, you know, very much exist. But, it's the ability to get the intelligence undertake the task in a very, very smart way to get the benefit at the end of it. So I think, you know, just having a GenTIC AI as as the end solution isn't isn't the right, you know, way to look at it. Patrice, you're nodding away. Yeah. I would no. I I I do agree with, with with John's approach, and I would kind of, probably have my three ideas, here about what to me is, is important. I think the responsible AI and trust, and of course it covers starting from traditional AI, agent AI, but in the lines of agents, it becomes even more, important. So guardrails, explainability, and compliance are to be non negotiable, in a regulated industry. So we have to make sure that, you know, companies are as as we do actually on our side are implementing, the right framework on responsible AI and make sure that you you stay in control in what you do. The second part is probably about contextual intelligence over raw automation, and we talked already a little bit on on that. So the ability to reason, plan, and act with context is probably what differentiates, Adjunctic AI from traditional capabilities. And the last piece is this human AI collaboration, because we talk about Agentic, but it's really about the, the agent boss model. So having employee orchestrating agents, maximizing productivity while maintaining oversight on it. So these are my kind of three, important topics, that, I think everybody should think about. Yeah. And I like that we've all mentioned human in the loop, human out the loop, where wherever the human is, somewhere in the loop. It is very important to have that element, and we will obviously discuss that slightly later on. My final question in this section is, can AI and AgenTic technologies enable banks to transition from traditional analog methods to a truly digital operating system? Robert, you're on my screen. Should we start with you? Absolutely. Thank you for the the the setup, Ella. I think you you framed the the the question with the word enable. And, yes, AI and AgenTic will power that transition. If if I sort of then twisted it a little bit, I said more pointedly, it's actually gonna force banks to transition to the other to a digital model. Because on the other end of any transaction, it won't take long. If you fast forward it in your mind, an agent will be actually interacting with the bank in less than a year to two years. So that's not a forecast. That's kind of a reality. If you look to The UAE, you'll sort of see true agentic transactions. People are already going on the chat TTP. They want to make a purchase. It won't be long before you step off the airplane and you sort of say, okay. I'm now traveling. Please tell all of my credit card companies that I'm now actually traveling. I don't wanna go log on to six different apps. So, every associate in the enterprise will either be assisted, augmented, or find their activities as I used the words before automated. And then they'll have to no associate can keep pace because on the other end is a digital, if they're not assisted. So if the bank doesn't keep pace, and they're not super powered inside of the organization, they stand no chance to face off. So I think they are not going to enable. They're going to force the banks to have to catch up. It's like you need to get on the train now before it leaves the station. And if it leaves the station, then it's gone. Or or you'll be running really damn fast. Right? Yeah. And you so, John, do you have anything else to add to this? Yeah. I think it's the catalyst, actually. I think, you know, typically, innovation is always applied to an analog operating model, and therefore, it doesn't reap the benefits. When you've got something like Adjunctic AI, this is the catalyst to get to a digital operating model, and you're already seeing it in some of the, you know, the, you know, the leaders in the market. I was just I was just with a client recently who've already deployed agents into trade finance as an example. So, you know, 30 agents in there undertaking a digital, you know, the scenario for letters of credit. So it's already live, and it's already, you know, operational, and you'll see that emerge quickly into other functions. So it's a catalyst ultimately to get to digital. So let's move on then from the why. We've talked a lot about the why to to the how, the practical steps that we can take. So what are some of the nonnegotiable foundations that are required to ensure a secure and sustainable scale? Robert, should we throw it to you? Thank you, Ella, for the the question. If you were visualizing this, you might see this in slabs or layers. As I sort of said, there's probably late afternoon hunger pangs forming. So I think of it as a six layer cake with the seventh layer being the icing. The first piece you've got to start with is design and approach, then you're gonna run into, ultimately, a data piece. A lot of people forget that data piece or having to realize they have to go back and reform that. You move up one layer from that, into the tech stack. How are we gonna tackle this? Then you move into that model risk and governance. We've got everything that we needed. We've got the model risk and governance. Then it's the operating model, which is how we're gonna run this on an ongoing where is the talent and expertise gonna come from? And then the last piece on top of this is really the icing, which is the AI governance and ethics because you have humans in the loop. You have machines making decisions that are actually making, I'll sort of say, impactful, or even life altering decisions. So you have to have that AI governance and ethics layer sitting on top. That's how, I guess, we would look at the secure sustainable scale in the foundation layers. Brilliant. And, obviously, it's Thanksgiving, so you're allowed to be hungry. Patrice, do you have anything to add to this? Yeah. Meaning, I I like what, what you are framing, Robert, about, your key points. And and starting with data, I I used to say there is no great AI without great data. So data is absolutely key. It doesn't mean that you have to, kind of move everything around and centralize all the data, but you have to have a data strategy, that is important. And data strategy comes then to, extend to cloud, actually. And and very often, you know, cloud is a is a foundation, because cloud, where cloud is by when I when I talk about cloud is about public cloud. It's really about economies of scale, economies of innovation, through agility, security, even sustainability that is important. But that's the tech layer and and Robert too, you elaborate about a couple of other things. So I think step number one is about strategies, strategy and execution. If you don't have a clear strategy and you need to have an AI strategy, in place and then capabilities to execute around that. The The second piece is about people, and and, you know, competencies, and and we might probably think about new roles coming in, companies. Tech is, of course, the given. And finally, no surprise, compliance and governance is absolutely key as you have to make sure that this new operating model, has has strong guard weights when it comes to, operated. So let's talk about that that human element, the new roles that we need, everyone's favorite subject. So what are these new roles, or the changing of existing roles, the executive teams must prioritize when hiring and training, in this new agentic world? John, should we start with you? Yeah. Absolutely. Thanks, Heather. It's really good to look at, you know, particular analogy. And if I go to governance risk and control, it's it's it's that perfect, you know, function where, an orchestration there could, really be beneficial. Because if you look at, the thousands of people who who sit in the GRC function at the moment, you know, their primary aim is to ensure they've got a risk framework, which is preventative. But in reality, you've got a very you've either got limited controls or you're reactive with a little bit of proactive elements there. So if you look at a a multi, orchestrated GenTIC, the model being deployed in, a GRC function, what you're gonna do is you're going to ultimately perform a whole series of functions. And maybe, literally, you you would have a a reduction of, workforce where where the actual human in the loop, it just becomes, you know, literally an observability element across a whole series of functions from first line to second line, and there's a third line, rather than being a deep specialist in liquidity risk or or FX risk. So you've got, the agentic orchestrated model performing, a whole series of tasks, but the human is really the observer and making sure that, there is a move to a much more preventative risk model as a result where you're focusing on sort of the key areas rather than being, you know, reactive. So the person becomes more of like an all rounder, basically babysitting all the Yeah. The agents who don't sleep, don't don't moan. Just get home a bit. Mic moan. No. No. They might moan. So, Robert, have you got anything to add that you just popped up on my screen? Yeah. Yeah. I mean, John's you know, when you sort of said the what new roles or whose role changes, I guess, in in in a word, it's like everybody's. Because the world's moved from coding to workloads to context. Context is all with the bank's existing teams. And then, you know, John's hinting not even hinting at it. He's sort of saying, you know, getting that whole enterprise and the whole organization means both organizational change management. To highlight two other roles, so, you know, John's sort of saying those specific roles, you know, there's a focus on registries. So you have to have people who are thinking about the risk mindset that John's describing. Then you have repositories where there's a group of people whose sole job is, okay, what we develop you John referenced, you know, if you have 30 agents, if you have 3,000 agents, that repository becomes super important. And then, obviously, you gotta have some type of part of the organization that's providing the tech stack foundation. And that that didn't even exist or doesn't exist. You know, that's each organization is spawning whether they call it an AIA team, a data analytics and AI team. You have to have that. And so those rules are proliferating in those small ones. Yeah. Ella, if I if I may maybe add one or two other roles that I might see from from our side. Absolutely. I think ROI analysts would probably, be also a key role, to look at, you know, the efficiency of of the model. And I was talking about, you know, having a strategy and execution piece. So AI strategist, might probably be also key when rethinking the way a company will operate in the in the future. Of course, data specialist, and and, and so on will be informed from from the tech side and finally AI trainers is kind of make sure that you build the right capability and also change the culture, in the organization. So these words might probably be also popping up here and there. So you've perfectly answered the next question. So I'll just move on to building a strategic pathway for scaling agentic transformation. So what strategic vectors allow banks to maximize ROI and accelerate enterprise wide adoption? I mean, Robert, if I throw to you? Yeah. Sure. I'd happy to thank you, Ella, for the you you know, when we talk about it as the the team and, you know, the the the four of us authored something around this. So so one of our colleagues is in in absentee. But we see kind of and talk about it around three vectors. So we've talked about a vector one, vector two, vector three. Vector one's about optimizing sort of what we sometimes refer as a great refactor, the great rewrite, and accelerating velocity on how the bank works today. So, you know, maybe it's SDLC, maybe it is other processes that just take too long to deliver. So that's a vector one when we say your second strategic vector, because this is the way you frame the question, is probably looking at AgenTic to drive new value and or address the new risks with AgenTic, kind of what you had also sort of said when do we move beyond that process efficiency. So that's our vector too. And then we would sort of talk about the third vector, and this is not a gross simplification. It's like vector three seeks to find optimization to fund the future. And that vector three then is is is sort of saying, okay. What processes do we have in the organization, not necessarily a software development life cycle, are just broken and would just get sort of, supplanted by dozens or hundreds of agents. And then, you know, that's the strategic vectors in Accelerate Enterprise. But then you have to sort of say, which one of those do I prioritize? Because if you say, oh, we're gonna do all three and we're all gonna prioritize all three, you probably will fail on what Patrice really you know, the the ROI, analyst stroke, leader is gonna come back and sort of say, congratulations. You did a lot. You ran in a big circle, and you ended up where you started from. You didn't generate ROI. No. You know, that doesn't work. So in short, declare your priorities in a mark. Pick one of those three vectors, then grab the other two and bring it along, but pick one. And, Patrice, you're nodding away. Yeah. Meaning, it's an interesting, conversation on this ROI, and I guess we we would touch probably on it later. But when you look into, these strategic pathways, you have to start somewhere and you have to demonstrate the capability to execute. So it's still about prioritizing high impact use cases, where you can immediately demonstrate that, you know, there is more savings, efficiency, innovation behind that. But we should not kind of stay to adjust use cases. Then it's about building a a scalable modular architecture, in order to kind of scale whatever you are you are you are learning to to execute. It's about responsible AI and compliance, this I mentioned already. It's about talent and change management, so it's it's tough with culture and and risk killing or risk killing whatever, you have. It's about partner ecosystem. No one can be successful alone because, you know, every enterprise organization for bank might operate in a in a broader ecosystem, and this is also our case as we serve this this industry. And it's it's done having a good understanding on how you, how you roll out, your solution in in a phased way because, as we say, in France, Rome didn't build up in one day. Mhmm. You will have to kind of look into what is the plan and how to to face your your strategy. So it's not everything, basically. But think about just if I can just add, Ella, just just in terms of you know, if you look at, traditional approach when, doing a big sort of cost, you know, reduction program at our clients, you typically go into the operating model and calculate the total cost of, the operating model with the TCO. And I suppose there's a major component part around, the Feet headcounts, you know, within those operating models. And the question we need to ask ourselves is that, you know, you tend to get, cost savings when there's reduction in Feet headcounts. You tend to get cost savings or revenue uplift when there's efficiency. And I think one thing that we certainly need to think about is what's the cost of an agent. So when you start deploying thousands of agents into into an operating model, we need to calculate that cost. Because if you've got an agent doing a single task or an agent doing a multitask, you know, is that actually adding efficiency or not? Because there's a lot of energy that's being used in order to run these. So the last thing you want to do is suddenly deploy thousands of agents and suddenly realize actually your total cost of ownership's gone north, but gone south. And, I think a key KPI is what's the total cost of an agent that needs to be factored in in order to see the ROI. A really good point, actually. Because I think people get so excited by the shiny new toys that they don't think about actually how to implement them. They just think, let's go and try all out. So how how must the traditional model change then to account for this new speed, agility, and this transformation that we're going through with Vergentic AI, Patrice? So I think we we still have to, as as I told you before, have a good understanding of, the savings or the efficiencies or the ROI TCO ROI are are important because, again, it proves the the case. Our learning when, you know, when we launched Copilot and, you know, the first kind of use cases, and and Agile, I think, is now the extension of, you know, Copilot being the the the interface and then agents executing behind the Copilot. When we started to look into ROI kind of conversation, you know, it can very often to saving minutes or hours, which is great, because it it gives you a sense of a, you know, shortening process or giving the opportunity to do something else. But are we willing to feed confirmation business, in in order to, have these minutes unused, or is it just for the sake of, you know, balance, a good balance between work and and and non work? So it was an interesting case, and and and, again, it needs to happen, in order to identify where are the savings, where potentially we can increase value, could be NPS, could be a reduction of churn, could be a call center efficiency metrics and so on and so forth. But what is also interesting is that, as we scale the model out, company wide, there is a moment where you can't measure everything. So there are pieces you have to measure and there are pieces where you you have, I would say, to give chance or to let the chance to people at, and and and and and deliver on their human ingenuity. Meaning, giving a tool, not knowing 100% how this tool will be used, but but trust people to be clever enough to make it happen and make it smart and make it faster, in order to save time, save energy, and create additional value. The world is full of these kind of examples where we had something that was bought in one purpose and was basically reused in different ways, in different shape or form that creates even much more value than what was expected. So net net, there is, of course, needs of models, to start calculate because it's not kind of a, you know, country. So you have to measure what you can measure, but also think about strategy, how to empower, people to use technology to do much more. Yeah. And I guess kind of also evolving with it. So trying it out, seeing how it goes, you know, altering what you need to give it to make to get the best value out of it. Exactly. So, Robert, I saw you took your mic off for you. So do you have something else to add on this? Yeah. I I only wanted to ensure that, you know, ROI is the value of benchmark and it's as valid as ever. So I don't think any of us are trying to say move away from it. Patrice's own comment about meeting someone who does just this for the AI. So we are all sort of saying it it's fine or bring in more money or lower the cost to produce. None of us suggest that you have to move away from it. It's just, at the moment, people probably have not even begun to scale this to a level that gives them the chance to generate that ROI yet. And all you know, I only wanna make sure that we're not I don't wanna think we're like JK Rowling, gnomes and bankers, but we're like, there's money or there's no money. There's ROI or there's no ROI. None of us are suggesting, I think, that we move away from that because your question said, how do you move away from the traditional measurements? Can't move away from it. Gotta measure it. If you get a little extra time, but you're gonna get measured on it. And if it's not producing, as John said, you can't tell me how much the LOM cost on that agent. You had a problem in less than a year's time. You need to be able to answer that. That was what was moving at me as I was listening to Patrice. Right? It's just like, don't none of us wanna say untether from ROI. We all think ROI matters. So as John was saying, obviously, cost, per agent is really important, and cost can be such a dominating theme when it comes to identifying the enterprise. How should leaders look at that, and are there efficient ways to deploy agents? If we start with you, John. So I think, there's gotta be some test bets, on this and that. And I I'll keep on going back to GRC. So I think GRC is a really good, you know, model to use because you're looking at securing the bank in terms of a whole risk framework. And I think the agentic, orchestration layer, you know, actually lends itself extremely well to this because what you're trying to do is you're trying to avoid fines. You're trying to, ensure that you're you're you're you're adhering to controls, and you've got a preventative culture, and you've got an observability. So I think it's a good element of, you know, you don't have to hire thousands of people to run, you know, the risk function in the bank because you've got that orchestrated layer. And it feels like you'll get this element of compliance will be done quicker, smarter. There will be a a genuine move to a preventative, you know, life cycle, you know, within the function, which means you don't get fined. And it feels to me that that that would be a really good example. So, you know, that's one. The other one is if you look at the cost per trade or the cost per loan or the cost per letter of credit, that would be a really another key example to say, right, if I can identify a post trade for a fixed income instrument or a or a derivative, then that to me then will reduce the cost increases the margin. So, again, it's a it's, you know, it's a usual nebarometer. So do it to avoid risk and and and and being fined, and then focus it on a on a tangible, like, cost per trade in order to get the benefit. Brilliant. So well, let's talk about the future now. So success in Agentic would require partnership and ecosystem collaboration. So what are your views on where that would be most advantageous for banks if we stay with you, John? Yeah. I think the partner ecosystem is fundamental to, you know, the benefits of, identifying the bank and, obviously, our very close relationship with Microsoft is a key element in Microsoft being a front runner in this, and a frontier organization. But there's other enterprise platform, their providers, in the mix, be it be it Oracle or be it the ServiceNow, who are also identifying with their platforms. And then you've got the innovators that are coming through like, you know, synthesizers, as an example, to speed up the SDLC or ADLC cycle. And you've got to knit those altogether to get the benefits. So you've got to get the infrastructure players right. You've got to get the hyperscalers. You've got to get the platform providers and then the innovators and even the fintechs. And the fintechs tend to keep the call, but it's important to make sure that you've got an end to end holistic, the solution going into the banks. And, Patrice, I feel like you've been nodding away. Yeah. Yeah. I couldn't agree more with what John said. You know, we are living in a world of ecosystems more and more. And and, you know, you need partnerships, and you need cooperation, coordination, orchestration, and so on and so forth. So to me, partnerships accelerate innovation and reduces risk, when when it comes to reshaping organization or helping to create additional value. So look at the the, the partnership between Microsoft and Cognizant, for example, and the way we could jointly bring value to, to to the market. And even thinking about competition and and John, you framed it pretty well, meaning everybody's identifying somehow their solution. So and there will be a layer of orchestration. And think about a very simple user interface, that will be natural language, for example, that that could even change the way we build software in the future. Why building, you know, software with a kind of predefined screens and so on and so forth? Tomorrow, it will be about interacting with data. So you will interact and you you ask against your data and and actually the system would build up the best appropriate, kind of a user interface coming with information out of different systems, different agents from different companies, even from different industries. If we think about the new upcoming feeder, which is, you know, sharing information across across the board, I think it's an exciting world and and partnership and ecosystem is absolutely key in order to be successfully building solution in that world. Do you also think, you know, collaboration in this space helps with, keeping up with regulations and making sure everyone's guardrails and the security is intact? Yes. Of course. And and look at, you know, in in Europe with DORA, for example, meaning DORA is, you know, the digital operational resiliency, act. And we have been recently designated as a critical third party, which, of course, for the industry should resonate very positively as we are part of the game now. And failing on our side with the platform, with the way we operate, technology by by by bringing us fines, right, or expose us to to fines. So it it becomes a really serious business in a already highly regulated industry. And and looking into these partnerships and and and cooperation, will help, kind of facilitate probably for banks, to build, systems, based on on trustful partnerships. Brilliant. And as Jonathan mentioned earlier, everyone is, identifying. So, how does AgenTic AI act as that acceleration engine for other emerging technologies like DLT and tokenization within within banks? If we stick with you, Patrice. So I take my crystal ball out when we look into the future and Yeah. All these kind of things. But I would say Convergent unlocks new models. So you're looking to AI and DFT, which enables secure, decentralized data sharing, and this will be important in the light of a feeder and and what what comes next. If you look into tokenization that creates probable assets, quantum computing, that will definitely revolutionize risk modeling and encryption. We think about, you know, kind of supporting that as a service, so building that on on on our core capabilities or platform that executes across across all these capabilities and expose this as a service. So, again, instead of having everybody try to develop by themselves things, we try to make it as standard as possible and expose that as a as a service. It it it will be a strategic imperative for banks as I think they would need to, probably explore hybrid strategies, AI for automation and personalization, DLT for trust and transparency, Quantum for, as I said, utmost security, portfolio optimization, and these kind of things. So we're preparing for the future, and by preparing for the future, you always have to think about, building the right foundation. And the right foundation is having the right tech stack, having the right partners that understand the stack and be able to build on top of that and then help, you know, kind of accelerate the movement. And, John, when you look into your crystal ball, what do you see? Well, I quite like, actually, Patricia's crystal ball. I think, it's pretty good. I'm a big believer in confluence. So confluence of innovation is key. And when we go back to the concept of ROI, if you get the confluence, you can then move the dial. You know, innovations like DLT took a little bit of a back seat, you know, when it first emerged five to seven years ago. But but the the actual relevance of it now might be become more important as you have an orchestrated, agentic, you know, model. Everything's auditable. You're suddenly gonna have a massive audit trail in terms of what every single agent is doing across the piece. The storage of that might be best to be used using a DLT rather than a in in a common database. So, you know, good use like that. The other element is, when we look at tokenization or, you know, quantum, it almost sort of, provides you a leapfrog moment in terms of where you can take this now because the ability and the reasoning and the context can come into play, you know, using agents, especially around trying to understand how best to use a number of technologies to make a big impact. So, I think the next, you know, ten years are are are years of renaissance. Right? You know, we're gonna have an enlightenment period, you know, by the time we get to 2030. By the time we get to 2035, you'll see the dial has completely shifted where we've got an autonomous world where, you know, the actually, the models are light and those innovations are are intertwined. Brilliant. So we're we're quickly running out of time. So I've got one more question. And if I stick with you, John, just to finish this off, if you were to be sharply critical about one thing you see today in the context of AI, what is it? I think we might not be spending enough time on r a RAI and SAI. RAI meaning responsible AI, and SAI is sustainable AI. So those are the two components that I think people are getting very excited about efficiency and autonomy, but we've gotta go back to some key fundamentals in terms of as we build the future of banking, making sure it's responsible, and making sure it's sustainable. Brilliant. And, yeah, like I said, we're running out of time. So I'm gonna give you guys thirty seconds on each of these questions. We've got some presubmitted questions, from our audience. So the first question is what immediate and long term actions should bank associates and leaders take to embrace an agentic workforce? Patrice, if I give you thirty seconds on that. Yeah. So I would I would go for, you know, my, agent boss model, and looking to our short, medium, and long term. Short term is probably human first. So every employee has an AI assistant or kind of a persona agent that helps them to work better, faster, agent, take over repetitive tasks like KYC compliance checks, loan process, you you name it. Right? Which which, of course, creates a a big, big traction. Midterm is about human and agents. So having teams, including AI agents managing managed by humans, to complete, tasks in an autonomous way. And when we think about maybe a little bit of long term, this is about agent and humans where humans set direction and agents will execute business process, checking in as needed, and and this is where the orchestration about multi agent workflows require risk killing of, AI literacy and data driven decision making will probably come into a place. So short, medium, long term, should be, well sequenced. Brilliant. And, Robert, if I end with you, what are the pros and cons of focusing on core system reinvention versus customer centric enhancement for the initial strategic pathway? It's a dozer of a question to answer in a few seconds. I'll take your challenge and sort of say if you know, I'll build on something from before. You know, we talked about VectorOne, and VectorOne is, in our minds, a necessity. That's, you know, how do you do the processes that you have today in AgenTek SDLC DevOps to drive velocity, quality, product impact? So, you know, what are the pros and cons? There's probably only a pro because you've reached the apex of efficiency in this. And I guess the con would say, you know, metaphorically, no one would say exercise has a contra other than the time you have to carve out to do it. So the only contra here is, you have to invest time in the foundational pieces. You have to invest that. So I think the pros and cons, you need to do that before you can even begin to touch either vector two, which is how do I touch the customer, or vector three, which is how do I make my efficient you know, my organization marginally more efficient, lower my cost income ratio? Brilliant. And I think that was probably about forty five seconds. So you smashed it. Well, that is unfortunately all we have time for. A massive thanks to our speakers and everyone who has joined us today. If you would like to learn more, please visit the Cognizant website for free blogs further detailing this topic. You can also find all three of these links on our landing page. This webinar will be online shortly, so don't forget to share. Thank you all and have a great afternoon. Bye bye for now.