 Live from Las Vegas, extracting the signal from the noise. It's theCUBE, covering IBM Insight 2015. Brought to you by IBM. Howdy folks, Dave Vellante and Paul Gillett. We're back to Las Vegas, everybody. This is theCUBE. We're here at IBM Insight 2015. Check out ibmgo.com. It's the digital social experience for IBM Insight. You can register so you can watch the keynotes. We heard Bob Picciano sort of hosting the keynotes this morning. Jay Poreway was sort of kicking things off. Both of those individuals are coming on. Tomorrow, Jay Pragg is here as the CIO of executive strategy and information at FNB South Africa, a major financial institution with a really interesting story. Jay, thanks for coming to theCUBE. It's good to see you. Thank you. So, we were talking off camera about what you guys are doing, but this is a great story of what we call systems of intelligence, bringing transactional systems and analytics systems together to this new modern world. And you're beginning that journey actually well into it. So, start with FNB Bank, what you guys are all about and what your role is. All right, so let's go back. I've been at the bank for 32 years, so let's just start briefly in terms of where we were 32 years ago. 32 years ago, we were then called Barclays Bank. We actually were Barclays Bank. And due to sanctions in South Africa, Barclays Bank pulled out and we then became FNB. And as part of that conversion from moving from the old ICL systems we had, we migrated on to IBM and a platform called the Hogan Banking Platform, Core Banking Platform. And at that stage, we made a very strategic decision to use IMS, DBDC as our core transactional database platform. And I think that decision was a very wise decision at that point in time. If you come to think of what we've achieved today in terms of IMS, the application designs underneath it, we have plus minus 30 million lines of code, which are continuing to evolve and sustain that technology as we move along to our 2025 strategy. So it has helped us a great deal, sticking to that technology and making it work for us. So a lot of people would say 30 million lines of code, oh, mainframe, IMS, why don't you get off that stuff? Why don't you get off that stuff? Many reasons, but I think the most important reason is the reason of cost. And the actual reliability and scalability of that platform. So if you look at, if you want to scale a IMS application and we are, our volumes are increasing and more people using our mobile platforms to scale is extremely simple. You just fire up additional CPUs, you fire up additional IMS servers and you're up and running with more volumes. So as long as you've designed your applications in the right way, you're able to scale, there's no limit to scalability in terms of IMS. So that's the first reason. The second reason is got to do with the cost factor and that is today a typical IMS transaction will cost us under a cent, that's a South African Rand cent. And that is extremely cheap to run a transaction performing all the complex functions of payments, deposits, withdrawals and some very complex corporate functionality as well. And you find that in terms of that cost, it's extremely cheap to run and therefore you're leveraging that cost into the business as well and therefore the business has become a lot more profitable because they're not incurring that huge cost of IT. So given the technology changes in the last 30 years, you've obviously assessed the alternatives and you've determined that it's the least expensive and then there's another factor too which is risk, isn't there? I mean if you had to convert all that COBOL code and get off the mainframe and go to whatever, pick your platform target, you could put yourselves out of business. 100% because once you do that, the business loses its insight into making money. So what they're focusing on is taking this huge monstrosity and trying to convert it to another platform which you're not even guaranteed. It's going to give you what you want out of it. People don't really understand that nuance. If you were to do that and I've had clients try to do this and if you have to do that, you have to freeze the code. If you don't freeze the code, you'll fail. I've seen clients try to do it without freezing the code. They've almost gone out of business. One in particular cost them $600 million to save 10 million a year. But if you freeze the code, you can't make changes to it so you can't stay competitive. So you're in this position where you've embraced it and said we're just going to keep adding value to this platform. And you must also understand from a bank perspective, much of the, because we're such an innovative bank, most of the innovation happens through our channel layer, our mobile platforms, our online banking layers, our branch platforms, contact centers and so on. And but they use the core, the core banking platform as the center of all transactions run through that particular core banking platform. So what we've done is we've made the life of a digital channel a lot more simpler because we use what we call a services approach. So you write a service once, it's available to any channel straight away. And that philosophy around designs, around architecting solutions the right way allows us to manage that growth of our products and services that we have as well. So it makes businesses a lot more agile. And yes, agile is actually probably a bad word to use. It's misunderstood, it's misconstrued. So you always end up saying that business will always tell us, but we're not agile, but you can see what we deliver. And because we're delivering so often so quickly, we're effectively a very agile environment. We were talking earlier about how you are exposing all of your functionality as services internally. Were there any special complexities using COBOL, IMS, using these older legacy platforms and creating a service-oriented architecture from that? So remember, we started with the service-oriented architecture designs of it in around about 1999, 2000. We deployed in 2000, I think. And at that stage, SOA was never matured. If you buy any SOA platform, it wasn't as mature as what they are today. So we've taken some of those learnings, taken the fact that we need to cater for scalability, to be as close to the IMS transaction as possible. And we decided, the decision we made then was to write a homegrown services layer that allows us to create the concept of services in COBOL to deploy onto channels for them to use. So we have a very interesting concept around our services architecture. It is our single point of contact for any channel, and it is a single point of contact to any product-house systems we have out there. So it forms the so-called integration framework between the channel and the core banking layers. It hides the complexity away from the channel. This evidently has not been a problem in terms of the bank innovating. You've been received rewards for your innovation. You had the first mobile app in South Africa, I understand. How do you continue to move forward with, as new technologies come online, how compatible has your legacy environment been with where you want to go? So I think to describe that, one needs to understand the business model we operate. The business model is one in which we operate a federated banking environment, which means we have multiple business units within FNB. And each business unit has its own CIL and CFL, and they're responsible for generating and focusing on their business strategy. And that creates the innovation around it, because you're focusing on your products, your set of services, and you want to try and manage new products and services and new innovative products through your business unit. And the key to this is not just providing it to your business unit, but also making sustainable and available for the rest of the business units to use. So we have a customer view of our customer. We can see exactly what products he's got, what services he's got enabled with, and at a customer level, we're able to do things very differently compared to not having that view. So for those of us who've been in the business a long time, you've seen a lot of changes, amazing changes, but at the same time, we will often comment, wow, it's almost the same thing all over again, except, you know, circa 2015. Thinking about transaction systems, what people now call systems of record, and then people said, okay, well now the things that matter are systems of engagement. That's two, three, four years ago was all the social media craze. Now it's IBM calls it systems of insight. Others, we use the term systems of intelligence. They pose these terms as though they're sort of unique bespoke, independent entities, but they're not. You're talking about an example where you're bringing systems of record or transaction systems and analytic systems together. Can you talk about that process and where you are on that journey? So I think let's go back to some of the regulatory environments that we were faced with. We launched just before the 2008, 2009 crisis, we launched what we call the National Credit Act. That implied that to get credit, you gotta be able to afford the credit. And therefore you cannot just approach one bank and the bank denies, oh, it does not want to give you the credit, you then go off to another bank and hopefully that bank will give it to you. So the National Credit Act implied that all banks have to subscribe to a credit program to say that they will only grant credit to people who can afford. In that particular environment, we had to make sure that as we matured in our technology, matured in our services, we were able to sell products through our online banking platform real-time. So I can score you, I can give you credit or real-time online. To do that, we needed the analytics behind it. So today it's not real-time in terms of the analytics side, but it's real-time in terms of decision-making. So in other words, here's a customer wanting to apply for a credit card and a overdraft facility for that matter. And we then score the customer, we look at his affordability, we get his credit records, we see if he was anywhere had any problems as a customer. And based on his records that we have, we were able to dynamically and at a real-time provide him that credit straight away. So he's able to transact on it immediately. And I think that is the type of analytics we played with over the last few years. We're getting to a stage where we're realizing that we actually need to move towards a more real-time approach. It's also around fraud and so on. So we needed to approach a more real-time approach and we're not there yet, but we're putting a lot of things in place to actually get to that way of transacting. That's one of the reasons this inside conference is actually very interesting because some of the things you can learn from here, you can apply and make it work in your environment in the way you would like it. So that's a great example. So I call it risk, generally scoring. The business benefit is you're compressing the time to the decision and you're obviously able to drive more business, presumably. And what about fraud? Is that a use case that is emerging? Are you there yet? So fraud is really big on things like plastic transactions. So in other words, credit card or debit card transactions. Over the last few years, we converted to EMV. And that has reduced the fraud levels. It still does exist, but by moving to an EMV card approach with the chip embedded on the actual plastic itself, we did limit fraud because now you require not just your PIN, but you require your PIN to be hardened to that particular chip as well. Therefore, making it a bit more difficult for the guy to manufacture a card. And therefore that created the drop the fraud levels quite nicely. They still fraud out there, but predominantly it's outside South Africa. So the credit card is reused in a country that does not play in the EMV world as yet. So some of the fraud elements actually happen in the USA because all you do is you go to a website, long as you know the guy's card number and he's three digit CBV, you're able to transact because there's no checking of the actual chip yet. EMV is not active yet in a big way. So therefore fraud comes from countries where EMV is not particularly well-employed. Yeah, it's kind of the Wild West here in the United States. It's true, I haven't been to South Africa in a long time, but even back in the 90s, it was pretty sophisticated from a banking standpoint relative to the US, but what is a, so how does the consumer transact business online in South Africa and what's the hard connection to the EMV, how does that work? Right, so most people at the bank, so let me explain a bit in terms of transactability. Most of all transactions are done through online banking and mobile channels and then obviously ATMs and point of sale. ATMs, all our ATMs are chip enabled and all our POS devices are chip enabled as well. So therefore, the chances of fraud are actually very limited in that way. So from a transactional perspective, we're using digital devices quite heavily. I mean, if I look at online banking, we do close to about 290 to 300 million transactions a month followed closely by the mobile devices. That's your smart app and your mobile banking environment and some of the prepaid things that we do in South Africa. Remember in a country where we home to a lot of disadvantaged people and therefore money has to change hands in a very different way, sometimes cash, sometimes digitally. So we're very big on the mobile space allowing transactions to be here, to be delivered through mobile technology. So example, I could pay you money through your cell phone via what we call geopayments and you could collect money at an ATM. So you don't have to be a bank customer to actually collect money from a bank customer. So it makes some of the things you've done with technology and in the environment we're in in the country makes better sense to use technology than make use of brick and mortar type branches. Getting back to your legacy environment and your commitment to your legacy environment. Have you had any concerns about compatibility of new types of applications you may bring in and their backward compatibility with the legacy infrastructure? And also where do you find coders, cobalt programmers these days? Right, so first things around compatibility. We run if the IBM mainframe obviously it's a full on IBM shop. We stick to all the maintenance levels that we have with new versions of software like the operating system, IMS itself. And we've recently purchased ODM as well which is one of the engines we used to manage our business rules with. Decisions actually done through the ODM engine real time through a cobalt API. So effectively we're sitting with an environment that is exceptionally robust. It is managed and maintained the right way. It is secure, we run the entire full stack of IBM cryptography on it as well. So it makes the environment very secure. And therefore to invest in another platform or try and do something different is going to be a very difficult decision to make. We can transition to something but at this point in time where we stand, the mainframe with all its setups makes it a lot more easier to manage. And then... And this is an issue that I think is often neglected in our focus on new technology that a company like IBM still has a lot of legacy technology out there. They need to support. They need to support you. How well has IBM done in continuing to keep you current and upgrade IMS with the new features that you see? So in terms of cobalt and so on, we've done very well in that space. And also the IMS setups in terms of where it was from the old versions. Remember I've been around 32 years. I've been around when IMS 1.2, 1.3 was around. If I look at the transition from 1.2, 1.3 to version 13 that we're running now and obviously version 14 that's just been announced as well, there's been so many changes across the board and the biggest changes has been support of large databases, support of open environments. So the ability to connect to other platforms in an open fashion. Using things like TCP IP, IMS connect, soap gateways. All those things that come with the product base now makes it easier to connect to platforms of other disparate platforms out there. So it's not as difficult as what people make it out to be. It's actually pretty simple to do that now. Game back to that cobalt question. Where do you find cobalt programmers for that? I was gonna answer that point just now. So what we've done with cobalt is over the last 10, 12 years, we started our own training academy. And what we do is we start the churning on between 15 to 20 cobalt programs a year. We've now up that limit to between 25 to 30 a year. So in about 10, 12 years, we've actually educated people of disparate cultures, people that never had the opportunity to work with computers, to extremely good cobalt programmers and IMS, DBAs and systems, MBS system programmers, et cetera, all coming through that academy. So and the academy is running for a very long while and with a very low attrition rate. So you don't outsource that? No, we allow scale development shop, we run, we do everything ourselves. Somebody at this conference last year, I think it was an executive from B of A, she said, the banking industry is starting to realize people need banking, but they may not need banks. So it was the disruptive theme, right? So your industry is getting disrupted. You're seeing Apple Pay and Square and all kinds of interesting things. And I was going to ask you about cryptocurrency. Where do you see those disruptors and how can you become the disruptor? So we've been pretty good at disrupting banking in South Africa, in terms of our innovations and so on and what we've done around banking as well. For example, we give you a reward for banking with us. So we pay out more in rewards than what you pay in service fees as a customer. And we reward you for good behavior. We reward you for using credit the right way. We reward you for making use of your credit card in the correct fashion. So as a customer, it makes sense for you to bank with someone who takes you as an individual, not just an individual, your family into account. We'll be able to work with you, because if you then go to someone else and says, you know what, bank with F&B because we believe that they have the best processes, the best banking platform around to help with all your needs, then that's a type of sort of marketing we can get out of people. So we treat people the right way. We wanna make sure we've got the right sets of customers and from there onwards, you will just continue growing. But disruption is there. There's a lot of disruption happening in that space around payments. And I think we are geared up to your point, but I think we're partnering with the right people. So I think we'll be fine. All right, Jay, we're out of time. But so I'll just, last question is, what are you learning at Insight? What are you getting out of the event? Do you have a good time? What, give us the bumper sticker on the show. I mean, I enjoy Vegas, it's nice. The nightlife's excellent, yeah. Just cause you don't come here like 20 times a year. Exactly, exactly. And I think what I want to get out of it is this whole real-time analytics at an IMS transactional level. And I think if we get that right, I think we are a long way to getting what we want real-time. Great story, Jay, it's so often it's not about the technology, about the people in the process around the technology and how you evolve that. So thanks very much for coming to theCUBE. It's great to meet you. Thank you, it's a pleasure. All right, keep right there. We'll be back with our next guest. As theCUBE were live from IBM Insight 2015, we'll be right back.