Sep 10, 2019
On this episode of COMMERCE NOW, we discuss how cloud-based banking is really booming and what FIs can do to be on the front side of this upswing.
Scott Anderson: Hello again. This is Scott Anderson, your host for this episode of COMMERCE NOW. As some of you have heard, cloud-based IT is booming, even with the core applications of the banks.
Scott Anderson: Although migration from conventional IT infrastructures to the cloud can be complex, once it's done, it offers interesting opportunities, not to mention enormous savings potential.
Scott Anderson: Today, I'm joined by Michael Engel, Managing Director and Vice President for banking software at Diebold Nixdorf. And on this episode of COMMERCE NOW, we will [00:00:30] discuss how cloud-based banking is really booming and what FIs can do to be on the front side of this upswing.
Scott Anderson: Welcome, Michael. Thanks for joining me today.
Michael Engel: Hey, it's a pleasure, Scott. Thanks for having me.
Scott Anderson: Before we dive in, Michael, I know you quite well, but why don't you begin by telling our audience a little bit about yourself and your background?
Michael Engel: Thank you, Scott. I'm in the industry for 20 plus years now and always worked in software and financial industry, [00:01:00] so I have seen a lot of iterations of technology. So all the way from the good solid host-based environments that a lot of FIs are still running, to now all these latest technologies driving banks and FI's around the globe.
Michael Engel: So therefore, I would say I have a good view of the historic aspect, [00:01:30] but also as part of my job, talking to customers around the world, and also to technology providers around the globe, have a fairly good understanding of the dynamics in different geographies and markets, towards what is still under discussion and what is basically real and what is in production today.
Scott Anderson: Excellent. [00:02:00] And really appreciate that you're joining us with that global point of view. And if we target our focus then to today's topic, can you kick us off and briefly discuss cloud-based software and why you think that's been growing in popularity lately?
Michael Engel: Well when we talk about cloud-based software, it is more than just putting a piece of functionality into a central located data center [00:02:30] that is accessible through standard protocols and is hosted by a central organization that gives basically, a cost advantage based on a dynamic usage model.
Michael Engel: That is what most people immediately have in their mind, but it's much more than that. It is all around technology around that, so the whole idea of containerization, the whole idea of APIs [00:03:00] to be used and consumed and the whole tools in order to create software, to test software, to deploy software to run and operate. And all of these pieces are now coming from the mobile phone.
Michael Engel: This is really, really cool and a bunch of nerds across the globe are looking into that. Into it has become [00:03:30] now, into a state of maturity where, now conservative organizations such as banks are seriously looking into that and the front runners are actually exploring that mission critical application, post development and operation and so showing extremely good results. And then that is what brought us to the point [00:04:00] that from being a buzzword and being the cool thing around the block, it has matured into, it is actually real enough and mature enough for a mission critical environment such as special services industry.
Scott Anderson: Interesting. So I know that there are several trailblazers out there globally and you've touched on a couple of their examples and aside from costs, what other [00:04:30] benefits, you know, if we have some financial institutions listening in today, how can you assure them that there are benefits beyond cost with cloud based software?
Michael Engel: Well, again, let's, let's fundamentally look at what is happening elsewhere in the industry. And then if you look at, for instance, companies such as Amazon, if you look into Amazon, not as a cloud service provider but as a retail organization, then by [00:05:00] and large the entire Amazon platform. So the shop that you see where you go and look for stuff to buy, but also the whole logistic engine, even engine behind that, everything is a very dynamic infrastructure. Why? Because Amazon has realized that they're living in a very fast changing environment and if they're not able to [00:05:30] adapt changes from the market, changes from customer behavior, integration of new products and services into their offering, then they're falling behind and they no longer stay relevant for their consumers. So as a result, they have worked intensively with other technology providers on a end to end, so called continuous improvement, continuous deployment, delivery chain, [00:06:00] tool chain.
Michael Engel: And as a result of that, they are basically creating an ongoing living organism of software that will evolve itself every give and take 12 months. So basically no code in there is legacy code. Everything gets constantly updated, enhanced, improved, and by that, change. Now that is exactly [00:06:30] how a lot of these smaller fintechs are today operating. They are not better or smarter, they are simply faster. And they are faster by utilizing latest technology in order to adapt to customer demands. And in order to automate servicing their customers, you do this by using software which is obvious. But only if your software infrastructure is able [00:07:00] to adapt the rate of change, then you have a real chance to stay ahead of the game. And that is what if you now think about the traditional bank, one of the biggest headaches that the CIO of the bank has, because by and large, the core functionality of the bank and there are parts of bank sells basically virtual parts, mortgages, [00:07:30] not tangible, consumer loans, not a tangible product.
Michael Engel: It is basically a digital product. Now if your core infrastructure that creates those products, that manages those products that sell these kinds of products, is sitting on a whole space infrastructure that's 30 years old, and it's very hard to change and very hard to maintain and basically your IT guys are telling you, "Well once or twice a year we can do a minor update [00:08:00] to that system but won't touch anything because it might break." Then this is not really exactly the agile environment that you are looking for in order to compete in a faster and faster moving market.
Michael Engel: So and that realization triggered for a lot of front running banks, the question of how do I get to a end to end tool chain that allows me in the same [00:08:30] way like Amazon does it or other tech company do it, how can we utilize really an end to end agile process of deployment? Because today a lot of these agile processes stop right in front of production. So you have people in an agile setup developing new functionality but then the host team says, "Well guys, very nice but two [00:09:00] releases a year. Thank you very much."
Michael Engel: Now how to get there. And then this is exactly where we have worked with major FIs to have a look, the tool chains and the technology from how you define functionality and how you develop functionality and how you test and how you deploy functionality.
Michael Engel: How to bring that into the context of your requirements of a FI with regards to stability, [00:09:30] reliability and also adding the context of prebuilt microservices functionality for a financial services organization. And that is basically the underlying driving force here. It is not so much predominantly to get the cost advantages of cloud deployment. It is getting into a much more agile way of creating new services for the bank's customers, [00:10:00] and then on top of that be able to deploy the new much more cost effective environment than before.
Scott Anderson: Hmm. So if I think about the conservatism of most financial institutions, at least historically speaking, certainly the CIOs are looking at this with an open mind. But if we, if we think about the paradigm of legacy processes like a C-Systems, you know, this demand to go agile yet [00:10:30] still have happening to be waterfall at the end delivery point. You know, how does cloud based software, how would we articulate this to the people who actually have to do the heavy lifting in the financial institutions? Is this, are we automating steps and making their life easier that have been manual in the past? How would you describe that to them?
Michael Engel: Well yeah, for a certain time you're actually doing things in parallel. So what you usually do is a step by step approach. You take a certain set of functionality [00:11:00] and you would move that off your existing, cost heavy infrastructure and put that into a much more lightweight or a cost effective production environment. And by the savings you are getting out of that you have so to speak freed up budget for new innovation investments.
Michael Engel: So, and when I say that it is, it is a process that takes some time and yes, you are at risk in a risk [00:11:30] environment where you need to keep running the bank within the existing infrastructure that you have inherited, while you're building up a new infrastructure that allows you to create in the CICD, the paradigm code and the creation of the code, the deployment of the code, the containerization of the code and then the hardware permission and software orchestration [00:12:00] is at the end of the day fully automated.
Michael Engel: So a lot of the tasks that have been very manual and therefore costly and therefore a lot of dependencies including all the test and integration work is not really end to end automated. So whenever today developers in our organization create some new code, every night that runs through automated processes, will automatically be packaged into containers, will automatically be provisioned [00:12:30] to the, to the respective part. We're using for instance tangible scripts for the software orchestration. We use open source technologies such as the need is, and then we deploy that really in an open source environment. And then a lot of this functionality that does work for you, in the past was highly proprietary, highly costly. And then in the meantime it is in its majority based on open source and that open source [00:13:00] because it's now used by millions and millions of people has reached a robustness level, a maturity level and gives you cost performance ratio that is really unseen before.
Michael Engel: And that makes it so exciting. So it has moved from, again, out of that nerdy corner into really mainstream availability and robustness. And this automation [00:13:30] then allows us to step by step, move more chunks on the existing whole space infrastructure over to this type of environment. And you get with every step that you take more and more cost advantages and gain more and more speed. And that is what FI's are now looking into. So what are some of the less mission critical things that I can get some early wins on? Can test the waters, I can get some experience with this [00:14:00] technology. And as it shows improvement in cost advantages, I can basically migrate more and more steps, chunks from my dynamic key infrastructure over to that cloud base environment. Again, the cloud at the end of the day is just the representation of the operational infrastructure. Behind there is much more of the, the adoption of open source technology, [00:14:30] the adoption of that CICD thought process and the adoption of tools and technologies.
Michael Engel: Just to give you an example, what I mean with that. So if we are addressing topics such as the infrastructure for managing payments within the bank where more of the core parts of the bank to the retail bank today. Then we usually do that in a very high available, high available fault tolerant [00:15:00] environment. And in order to do that, you usually have effective deployments of those solutions. And as of today, if you want to run really even active/active mode, you need to synchronize data to keep those two sides that run active/active in synchronization. Now there is proprietary, very good software available to do that. But it's also very costly. And as of today we [00:15:30] are able to do this really at a fraction of the cost with open source tools and technology. So we're using, for instance, for the data replication between sides an open source technology called Kafka, which is performing extremely well.
Michael Engel: Even in the largest installations with the largest data loads that we are working on, even those top tier FI's. So just a little example of in the past, [00:16:00] very specialized proprietary software, very costly, today can be easily replaced by open source technology. However you still need to have the knowledge on how to utilize that technology in order to ensure data integrity and then synchronize data across sites. But again, just a little example of step-by-step replacing traditional very high costly software infrastructure by open source [00:16:30] technology in the cloud. We have done this now with a bunch of FI's globally, successfully and the results are amazing.
Scott Anderson: So if I have my chief operating hat on or chief information officer hat on, you're convincing me of the benefits, but I'm a little reluctant to buy into how easy this conversion would be. Can you talk about what typical conversion processes look like to go from my traditional heavy data center operation to a cloud based operation?
Michael Engel: [00:17:00] Don't get me wrong. I'm not saying it's easy, I'm saying it's doable. But it takes knowledge, it takes investment and it takes some time in order to do that and the majority of the focus should be really in understanding, managing and operating that true change. And then that is much stronger as it is doable, but you need to have a set of good people that can help you [00:17:30] with that. And you need to have also a strategy on how to do it in the step by step migration because again, at the end of the day you still need to run a bank. However, the results and the benefits you are getting out of that are so immense, that tt is really worthwhile your. But the caution that I'm taking here is, it's not like snapping a finger and you're there. It is new technology, [00:18:00] it's mature technology, but it is also still a complex thing. If you have some help and you do have some guidance it is really manageable, and again the results are extremely beneficial.
Scott Anderson: So when you keyed on how this can be considered somewhat newer technology and I guess the risk adversity of, of some financial institutions, they have more of a wait and see mentality or I'll, I'll follow the leader, see how they do [00:18:30] you know, what is your view on how cloud based products have become better or more reliable over time?
Michael Engel: Well again the, the reliability comes through the mass usage because every system is only as good as it is fault tolerant, and how many hours went into usage and finding of bugs and eliminating bugs. And that is to a large extent [00:19:00] the advantage here of the open source technology community and also the cloud based technology. Because of the wide usage, all the cloud technology in production today and also the open source technology and tools around that, due to the fact that over the years now millions and millions have used it, it has reached the maturity level that before only [00:19:30] very expensive, proprietary technology could give you.
Michael Engel: So you get so to speak, the maturity level and the robustness by the multiplication of usage across the globe and you inherit those benefits out of all the endless hours that went in there. But the same goes for topics such as security because we see that in a lot of cyber-attacks today, that guy sits in the inside, so our systems are hacked with inside know-how and there is only so much security that you can prevent from the inside and in a regular bank, you usually have a team that handles this. This [00:20:30] is a number of 10 people, 20 people, and maybe 100 people. If you look at what organizations such as Amazon are doing, they have 10 times the amount of specialists, engineers and technology in place because of sensitivity in order to protect this type of cyber- attacks.
Michael Engel: And again in that sense there are sitting men on the outside side versus we usually [00:21:00] have half what we saw over the last years mainly happening. People who are inside the axis attacking problems with those kinds of systems.
Scott Anderson: So just touching on, you did hit the elephant in the room. Security certainly is of major interest I think to most FIs and looking at you know, all the publicity around data breaches and what have you. Just a little bit deeper on that Michael. So [00:21:30] if a bank is looking at this, what are the considerations for public cloud versus private cloud versus maybe a hybrid approach? Are there differences they should be thinking about?
Michael Engel: At the end of the day, it will always be an individual decision. And it has also to do with regional legislation and certification processes, things like that. So, you have certain countries where it is very clear about the legislator that certain data needs to stay in the country [00:22:00] for various reasons. So there are constraints that will decide whether you go public or private cloud and there are pros and cons on both of that. So I'm not saying that one or the other is the better choice. But again the decision to go public or private cloud again is only looking at the end state of deployment.
Michael Engel: If we consider what I said before, the cloud [00:22:30] is just a final representation of a fundamental shift on how you develop software, how you use tools, how you use an end to end agile approach to create software that will lead your organization into an agile state of mind that lets you stay ahead with your competition. Then the final decision, [00:23:00] whether you want that on your own data center or that sits on to somewhere else. It's actually a minor, is a minor decision and again, the needs we look at on an individual basis, what the cost advantages versus the certification slash legislation complications are on a case by case basis. [00:23:30] We have brilliant examples for those, we have very large FIs that who operate today, operate only in the public cloud.
Michael Engel: We have other customers that run on private cloud or in regional cloud type of environment. Either way is fine. The main benefits comes more from the adoption of a fundamental paradigm change, paradigm shift.
Scott Anderson: So with that I [00:24:00] think let's, let's come back to cost reduction. I know we've talked a lot about the other benefits around this, but if we look at the cost of this, what kind of cost reduction volume are we talking about here roughly? Is there an ROI that can be realized in a relatively quick manner? Thinking about this for a, I'll call it a large regional or a national bank, what does that look like?
Michael Engel: Well, what can you see it, and again, every case has, we looked at individually, but you can take certain data in terms of [00:24:30] how much transaction volume do you have, how much processing capacity do you need? And you can look at your, your legacy environment and there are formulas on how to project that into a cloud environment. But, a rule of thumb, we have seen across the globe a cost reduction for the operational environment, so the platform in terms of MIPS and database [00:25:00] and just space and everything else around always North of 90% cost reduction towards a traditional mainframe based operational environment.
Scott Anderson: So quite doable. Something that absolutely needs to be on the radar for people to look at this.
Michael Engel: Oh yeah. We had an FI that went recently, even from a host space environment to a already lightweight [00:25:30] cost effective Linux server environment for their production. However, and this is a very large FI running tens of thousands of branches and ATMs and whatsoever. They're current lightweight operational infrastructure right now runs on annual $10 million bill. If we would have put that into a public cloud infrastructure, and we haven't even [00:26:00] calculated the sort of speed that having adoption of utilization into that, but that would go down from roughly 10 million to below 400K just to do ...
Scott Anderson: Per year?
Michael Engel: Per year.
Scott Anderson: Wow.
Michael Engel: Just in order of magnitude. From the operational end.
Scott Anderson: So you've touched on seeing examples globally. I'm curious where are you seeing the biggest and fastest move to cloud based banking? Is it a phenomena in certain countries? What's driving that and what can others expect [00:26:30] and when's it going to hit them?
Michael Engel: What you see is this is a global phenomenon, and you have the technology savvy banks or the first movers already deep inside that process of adoption. They have been so to speak playing around with this since years already. So they have got their feet wet on the whole agile development, the all open source technology [00:27:00] topic, the whole open banking API idea. And now the deployment into the cloud is, so to speak with the final step of putting everything on a new, much more agile, creation of design creation, deployment and operating infrastructure. And yeah, we are talking to banks about that in Australia and Asia, and Europe [00:27:30] and the Americas so it is basically everywhere. And you have much more of the traditional early adopters, fast followers, slow movers type of scenarios in each country. But there is not a particular and geo that is faster than the other.
Michael Engel: You do see, however, in some countries extreme fast moves because they're coming so to speak from behind. They're catching up and they're [00:28:00] just jumping over a couple of iteration steps in, in technology and go directly to such a quote and quote cloud based type of strategy.
Scott Anderson: Interesting.
Michael Engel: And by the way, when we say cloud, I always mean a, what some people refer to a cloud native strategy, so it's much more than take your existing application and just apply it somewhere at the central data [00:28:30] center, it is really that end to end idea of how to create, in an agile manner, code that is really designed to be and take advantage of the capabilities of an agile, also agile environment, of a cloud operation infrastructure.
Scott Anderson: Oh that's a good point. I think so it's not just virtualization of a piece of software, it's the whole end to end process around it.
Michael Engel: Right. I mean in a lot of cases [00:29:00] the just virtualization of an existing application is a good step to get familiar with the subject matter and get some experience but it doesn't change the dynamics and again, the key change that we see here in the dynamics is how do I transform my organization to be much more nimble and agile in responding to market changes [00:29:30] and customer demands. And again, if my production environment that is in its core as it says, 30 years old and I still need to send a program out to a programmer to accept my demands, I am in a tough spot.
Michael Engel: And that is the dilemma that a lot of banks are in, because they also know that just changing the core banking system [00:30:00] is a very difficult task and cost most of the CIOs their career. How do we get there, how do we get there? And that is where we see to really create a step by step approach that we build up in parallel to the existing infrastructure and by reallocating existing processing capabilities and capacities to cloud, daily cost advantages that allows you to finance and [00:30:30] process over time.
Scott Anderson: Michael, I really appreciate the inputs here. I think it's provided some clarity and hopefully given some food for thought to folks who have heard about this, but not really thought much around how it would impact them and how they could implement. I think at this point, this is a great place to wrap up the conversation. We could talk for hours on this, but thank you again Michael for joining us and thank you to all of our listeners for joining us today. To learn more about topics like these, [00:31:00] log on to Diebold Nixdorf.com or click on the link in the podcast show notes below. Until next time, keep checking back on iTunes or however you listen to our podcasts for new topics on COMMERCE NOW.