Listen: Temenos on helping banks overcome AI complexities

Sixty-two percent of banks say the complexity and risks associated with handling personal data for artificial intelligence (AI) often outweigh the benefits to customer experience. That’s according to The Economist Intelligence Unit survey sponsored by core provider Temenos. While data can be a challenge, there are three steps organizations can take to better manage that data, Sue Laws, executive vice president of Business Solutions for the Americas at Temenos, tells Bank Automation News in this episode of “The Buzz.” “There is complexity, and there is risk associated with AI,” Laws says. “But once you take care of that, and then have the right resources to know how to best access and then leverage that data — that very much helps manage the associated risks.” Bank leaders should also have a clear strategy for AI, she explains. "When executives take a step back and look across the organization and see where their biggest pain points are from being cost effective, we need to have more automation in order for us to get better scale,” Laws said. “That’s where the executives are seeing where AI could best leverage, because AI can be deployed in so many different ways.” Listen as Laws outlines in this podcast explainable AI and composable banking, which is an agile approach to designing and deploying banking capabilities. Subscribe to The Buzz Podcast on iTunes , Spotify , Google podcast , or download the episode .

The following is a transcript generated by AI technology that has been lightly edited but still contains errors. Loraine Lawson: Good day and welcome to the Buzz, a Bank Automation News podcast. I’m Loraine Lawson, and recently I spoke with Sue Laws, the executive vice president of Business Solutions for the Americas at Temenos. The core provider recently sponsored a report on artificial intelligence projects at financial institutions. We talked about the survey’s finding that 62% of banks said the complexity and risks associated with handling personal data for AI often outweigh the benefits to customer experience. I asked Ms. Laws what Temenos offers in terms of support for AI and cores can do to address this issue. Sue Laws We not only offer an AI solution, but our AI is explainable. And explainable is a very important part of the offering. Because typically, it's explainability, that is incredibly important to the regulator. Because once the decision has been made in the event of somebody coming back to say, hey, we're not convinced of that decision, or maybe there was some bias involved. The ability to explain why that decision was made, is is the only way to have that full transparency. So yes, we do have a solution. It runs with our own core banking applications and solutions. But it's also very much designed to work independent of terminal solutions toLoraine Lawson so do you get that if you buy the core solution? Or is it a separate add on? How does that work? Sue Laws It's a separate add on. But in the world where we live in today, now, composable, banking, you know, any client that came to Temenos, we would be enabling them very much to compose what the banking services they would like to be utilizing from us. And then we would build brick by brick, as they say what that end banking services that are that they would wish to be offering, which could include Explainable AI, all depending upon what their needs are. But if some if somebody just wanted to buy a standalone maybe to put on a wealth platform that they've already got, we could be readily leveraging our Explainable AI capabilities with their, say, their well solution and their data lake or their data hub. Loraine Lawson So you said composable. Thinking what I have not heard that term, what does that mean? Sue Laws Composable banking is really a new agile approach and a new way in which we can be designing and delivering our banking capabilities. So that any one of our clients can, in effect, go to the portal, and self serve and pick and choose what capabilities they wish to be deploying. And then we can be delivering that to them for them to start readily testing and building products against it. Loraine Lawson Can you give me some sample? Sue Laws Yes. So in the days gone by when you would go for a monolith, you would maybe go to a bank and surprise to a software vendor and say, hey, I want to take your, your deposits and your lending capability. And you would just give them and deploy them the entire monolith of everything that's in that core banking application. And all you would be doing is turning on or configuring the products and the users and the capabilities that you would wish to use. But that entire monolith of an application is what's being delivered to you. When you go down to the world of composable banking, you're in effect taking bite-sized chunks of banking capabilities that you truly wish to consume. So you're only consuming what you need, which then also ensures that you are only paying for what you get paid for what you use. And it also made you a lot more agile, because you're just taking what you need to get to it's a faster and more agile way of deploying, designing and delivering banking capabilities to the clients. I think another way maybe to describe it in the rain, which I like is that financial institutions and banking solutions of the future are assembled and not built, because built suggests that it's going to be you know, like a big house and the roof sits on the house and everything. Whereas if we assemble it, you're enabling your clients and your customers to bring together those best of breed capabilities. Loraine Lawson So some people call that I guess microservices - is that? Sue Laws Microservices is one of the underlying architectures actually. Absolutely. Because microservices is a discrete business process that has been encapsulated. And then it's how you bring those different micro services together that you can then we starting to package up the banking capabilities. And each microservice will have an API, which has a means and a hook a means of talking with each other. And so it's basically building all those components together. And if you asked me to go to much more technical than that, I'll have to go. I need a propeller head to ride shotgun with me. But no, keeping it very, very simple, it just means that we take Lego blocks to build absolutely just what we need, when we need it. And then in the event that your volumes increase, you can scale up just that business process that you need to you're not having to scale up for the entire application. Does that make sense? . Loraine Lawson Yeah, it does. I might come back to that. But I did want to ask you looking at the survey, one thing that surprised me was that most banks agree --62% -- that the complexity and complexity and risk associated with handling personal data for AI projects often outweigh the benefits to customer experience. What did you make of that? And how can cores help with that. Sue Laws So you know, the complexity and risks associated with data is a very serious, it's a very serious statement, when we're looking at data and what what needs to be done with it. I mean, first of all, there's the responsibility to ensure that all the data is in one central place, because the majority of banks and certainly the bigger banks, they're going to have data in multiple places, they maybe have multiple cores, they'll maybe have multiple Master Data Management files, they will have third party systems, as well as the insurance, credit cards, debit card information stored in many places. So the first thing we need to do is to be bringing all that data together in one central place, be it in sorry, be it in a data warehouse or be in a data lake. Secondly, then on top of that, we need to be very much bringing together a standard policy. And that is to bring a standard policy of how that data is going to be accessed. Because obviously, there's a lot of as you say, there's a lot of data, there's a lot of privacy associated with that data. And then the third step that's very, very important is bringing together the infrastructure and the tooling. So that you can start to be writing and accessing the data to write the algorithms, and then the machine learning together to take that data and do meaningful things with it, and how to read the data and then how to use it. And those are three very fundamental steps, which need to be taken as part of embarking on an AI journey. And that's where, for me, it comes back to, there is complexity, and there is risk associated with AI. But once you take care of that, and then have the right resources, to know how to best access and then leverage that data that very much helps manage the associated risks that goes goes with that. But the risk appetite has to always stay with the bank and the compliance in their own own policies within as to how they want to best manage that. Because the other thing that I always say when you start talking about data and data privacy, and how banks and financial institutions take care of that you cannot put a value against reputational risk of doing something not right with that data. Loraine Lawson You mentioned your AI capabilities. Uh, can you talk a little bit more in depth about that, like in terms of how can that be used? How can that be leveraged? Where did it makes typically use for the AI? Is it just the capability and then you treat it as you want to? Or is it already trained for certain things? Sue Laws So are Explainable AI, I mean, we find the biggest use case for it is for fraud, you know, for monitoring, monitoring fraudulent transactions against, you know, like financial crime mitigation. So, the ability to be able to to look at all the transactions as that and the payments as they're going through the financial institution. By leveraging the AI we can be very much getting focused on reducing the number of false positives that are typically found from financial crime mitigation software, the helps to reduce those number. And by reducing the number of false positive, that also means the number of payments that are correctly straight through processing out the door increases, which of course then gives the benefits to your underlying customers. And just to explain what a false positive is, within financial crime mitigation software, there are rules to determine whether a payment could potentially look suspicious and require a secondary look at it, which without AI would mean would go to a compliance officers queue for them to take a look. And then they would have to make that manual determination and decisioning of you know what I know this payments, good. I know the underlying customer. It's not on a sanction list or anything like that I'm okay to release it. The AI and machine learning enables those false positives to be automated and go away. So better experience for the customer and customers and which is even more important if you're corporate with high volume payments going through the organization. And of course, it means there's less manual intervention, less manual intervention means it's, it's cheaper for the bank to be to be processing. And ultimately, we want to increase our efficiency, keep down losses and keep down the cost overhead costs. Our Explainable AI, let me back up to your original question of how long does it deploy, we have pre configured models, which comes with our software, which is obviously an accelerator for anyone we've pre configured models, we can be looking at deploying them with other software. But of course, we've also got the ability with the right technical resources to be creating our own algorithm to leverage that capability as much as any as anyone desires. So a data scientist would have a lot of fun with, with our, our Explainable AI. Loraine Lawson And it was that something you built in house or to acquire that or, Oh, Sue Laws we acquired the Explainable AI from a company called Logical Glue. And we acquired them in the July of what we are we now July 2019, if my memory serves me, right. And yes, it was an acquisition. And it was a very purposeful acquisition, because there were a lot of AI and machine learning on the market. But that explainability, which is very much patented, was one of the major reasons that we acquired that capability. Loraine Lawson Can you break that down a little bit what you mean by explainability? I understand that it has to do with being able to explain how it reaches conclusions, basically. But what does that look like? How does that happened? Sue Laws Okay, good question. So, if we think about a couple of years ago, there was an incident with Apple cards, where they became subject to an inquiry, because it was perceived there was bias or gender bias, based upon credit scores and the number of of decisions that were in favor of as diverse as down. They were unable to explain how the algorithm has come to that conclusion. From our perspective, when we look at the algorithm that's made a decision, it will actually be printing out in a readable format, where you don't need to be a data scientist to understand it from the red to the green, what that score was, and the underlying base simple questions, then, in effect gave rise to that score. So you'd also be able to if you're wanting to talk to someone where the credit score or the scoring didn't serve them, well, you could say in order to increase your score, so that you would pass next time, this is where you need to be improving. So our Explainable AI needs you to drill into the underlying decision, the underlying questions that are underneath and how that score came about. And then it could obviously be made available to a regulator if they needed to, to get access to it. And also it can be made available to anyone within the financial institution, to as I say, was working with a client that maybe is wanting to get a mortgage and wants to improve their score. Well, hey, this is how if you improve your score, this is how we can help you get that rate down. Loraine Lawson So as a use case for the to do alternative credit scoring, is that something you can support? Sue Laws That's that's another use case we have absolutely along with. We've got other use cases would include from a wealth management perspective. There's a lot of interest these days into that ESG aspect. of where maybe you want to be. If you want to make sure that your assets, your monies are being invested in companies that are very conscious about their carbon footprint, for example, that's a use case where the AI can be data mining to look for the companies that are proactively managing their carbon footprints and therefore, then make the recommendation that this would be an investment that would meet the needs of Lorraine, because she's very keen to only investing in companies that are carbon footprint conscious. Loraine Lawson The survey also found, well, it references reference a survey the report did about it, executives that found 85% have a clear strategy for adopting AI in the development of new products and services. I wondered, what have you seen? What does that mean to have a clear strategy? In your opinion? Sue Laws I think when executives take a look back across there and take a step back and look across the organization and see where their biggest pain points are from be at a cost effective, be a we need to have more automation in order for us to get better scale. I think that's where the executives are seeing where AI could best leverage because AI can be deployed in so many different ways from a very simple Chatbot. To take away just those those rudimentary touch points with a client, but the moment the Chatbot, the person says, Yes, that's what I would like to then get them engaged with a human to make sure that you keep that hands on touch, I see so much executives knows that they have a clear strategy that AI is important to them. But when we then have to deal with that next click down and then say, and how exactly are we going to go about that together? In the world today, where everyone's pinching on margins, you know, with the interest rates still being incredibly low, we really have to find a way to keep our costs down to increase our revenue, and to leverage the human resources in a capacity where they're truly adding value. So that we can leverage technology and in particular AI and straight through processing to take that more mundane and humdrum work out of that cost for the for the for the for the organization's Loraine Lawson okay, what was the most interesting finding for you from the report? Sue Laws Um, for me from the report, you know, what, I think it's not a Yeah, one of the most interesting to me is really just how simple yet efficient ultimately, AI is. Because banks and financial institutions, we all have access to data for years. And I mean, for years, the data has always been there. And yet, it's only when you really start to not just get access to it, but get access to it to do something meaningful with it. It's very easy to forget how important using that data wisely is, and how then we can be staying relevant with our customers. And because of the way in which the world of banking has changed so dramatically, nevermind over the last two years for COVID. But it's significantly increased exponentially over the last 345 years. But consumers like ourselves have become very much less loyal to the financial institutions that we use and have been with much more easy for us to get access to other banking capabilities by organizations offering them to us at the most relevant time of the transaction. And anyone you know, an example I'll give you it, it never ceases to amaze me how I might have been googling on my phone looking for something super exciting, like a steam cleaner cleaner. And then suddenly, I am met by that link in maybe a Facebook or something on LinkedIn comes in as an advertisement or even an email. And it's not just taking you to look at that thing cleaner. It's also telling you where it's available to buy in three other places and here's the deal. There's different prices. But it's also been saying in the event that I go and buy it with you. Would you now like to do buy now pay later or would you just like a short term installment loan? It's making it so easy for us to bank. Of course, there's pros and cons for that as well. But um, and that to me, though, is all done by the power of data. Loraine Lawson: You’ve been listening to the Buzz, a Bank Automation News podcast. Thank you for your time, and be sure to visit us at Bank Automation News.com for more automation news. You can also follow us on Twitter and LinkedIn. Please don’t hesitate to rate this podcast on your podcast platform of choice.

Listen: Temenos on helping banks overcome AI complexities

Why banks need a ‘clear strategy’ for artificial intelligence

Fintech Funding: Crypto exchange Binance.US secures $200M in first funding round

Data will be the key to innovating finance in 2022

Related Posts

Product roadmap, data governance key for AI vendor selection

How U.S. Bank measures ROI of AI

Acquisitions bring AI consulting in-house

Data will be the key to innovating finance in 2022

EMERGING FINTECH DIRECTORY

FinAi Podcast

SPONSORED

The Next Generation of Cyberthreats: Powered by AI

Build an Antifragile Strategy to Outperform the Market

How AI and Product Experts Turn Fuzzy Requirements Into Focused Dev-ready Roadmaps

Connect

Welcome Back!

Retrieve your password

Continue Reading with FinAi News Premium - Less than $2/Day

Listen: Temenos on helping banks overcome AI complexities

Why banks need a ‘clear strategy’ for artificial intelligence

Fintech Funding: Crypto exchange Binance.US secures $200M in first funding round

Data will be the key to innovating finance in 2022

Related Posts

FinAi Podcast

Connect

Welcome Back!

Retrieve your password

Unlock This Article

Continue Reading with FinAi News Premium - Less than $2/Day