 Okay, you guys ready? Yeah. Can you hear me? Okay. All right, very good. Well, good evening and welcome. This is a panel discussion, the title of which is Hybrid Cloud in Finance. And I've been able to, we've been able to get a selection of people from finance companies and related finance companies to join us with the panel this afternoon. And I think we'll have a pretty interesting discussion because obviously everybody in finance and everybody on any enterprise is dealing with a lot of really hard issues these days in moving to the cloud. So I'm not gonna spend a lot of time introducing each of these speakers individually. Their photo and their name is up on the slide here, hopefully in the order, almost in the order everybody's in. But I'll just run through everybody briefly. There you go. There you go, thanks David. So we've got Asif Alam from Reuters, Global Business Director from Reuters. Janga Alaminati from Visa, Chief Architect at Visa. Marco Hochstrasser from Swiss Carm, who does a lot of work with all the financials, or many of them, head of cloud platform dev there. My colleague David Lewis from BNY Mellon head of enterprise architecture practices there. And Andre Ribka from Bloomberg. So these are the people that are dealing with all the decisions and all the challenges and problems that cloud is bringing and the software delivery and all that is bringing to this industry. So it's a good panel to discuss some of these issues. And we're just gonna dive right into it. We're having some logistical issues with microphones, but I think it should work out okay. Yeah, we turned it into a stand up. Yeah, we'll turn it into a stand up. This is a stand up meeting. It will be passing the mics back and forth. Very good. So I'll invite you to just answer at will. You don't have to go in any particular order, but I'm gonna keep everybody up to three minutes or so on responses because we're short on time. So let's get started. So Bill Hilf this morning did mention a little bit about private versus public cloud and that whole concept is kind of evolving a lot, maybe in five years there won't be such a thing. But I've been interested in private versus public for years and the challenges that are involved in moving one way or other or the restrictions or constraints that prevent moving from one place to another. And I would love to get everybody here's opinion on that whole issue of moving to private, moving to public, what prevents you or forces you to go one to the other. And what is your concept of private versus public? Who'd like to start? I can start, I have the mic. So thanks a lot. So my view is that there was a concept of multi cloud which has been talked about a lot in the last two days. And I feel that that's where at least I feel the financial services is heading to. There will always be a place for public cloud for various reasons, entitlements, this, that, all the proprietary data that you have that you cannot move from one place or the other. But in the larger scheme of things, I feel that a multi cloud in public cloud with multi vendors, maybe even more than two vendors, is the momentum, if you look at the momentum or direction, that's what I'm seeing. Now, is it there yet right now? No, because of a lot of complexities. But that's where I see the industry driving towards and building solutions around that. Great. So in my view, at Visa, I mean, we chose private cloud. So it depends like, generally, like if the customer has a talent to build a cloud and what kind of infrastructure they have. If it is a new consumer application company and moving to public cloud without having any infrastructure or talent, that makes sense. But in my company, at Visa, we have a talent. We are running our data center for years and we have infrastructure. So we have a talent, we have infrastructure. Going to public cloud, paying somebody doesn't make sense. It's not cost effective to us. Building private cloud, initially it may be, I mean, challenging, but once we build the private cloud on our own premise, and so cost point of view, it's very, very meaningful. And also security, it's one of the concerns that we had. So we just, it's not about our customers, but it's about our customers' customers, customers, customers, customers. So we have to ensure the trust that's only possible with the private cloud. So that's the reason, I mean, in my view, it makes sense to go in with the private cloud. So you've talked about the financial reasons and the security reasons. Are any of those forcing you not to go to public cloud or is it just that's the sort of the least path? So if you have applications, I mean, which is not considered as a secured applications or customer dependent applications, definitely we can use public cloud. So it's like somebody's hosting it, we can just subscribe it instead of building and hosting on premise. Okay, all right, very good. Okay, Marco, to you. So we host a lot of financial companies, but I'm also at the foundation, so that's why I represent a couple of them. And what we see, especially in Europe, that the audits and compliance reasons are one important reason that the banks or financial services in general need to host some stuff on-prem still. And for them, multicloud is not just about having on-prem and public offerings and moving them around, but also have the possibility to, within their private environments, move it from the different clouds they have. Most of the financial service we see build in the last five years, maybe four to six clouds by their own and just moving workload around them and leveraging them best is what they understand on multicloud. Yeah, so a couple different perspectives, John, so you're probably fairly familiar with this. So I think we have a substantial investment in private cloud historically. We've built utility computing platforms for a long time. We own our own data centers. So as compelling as some of the financials claim to be for some of the public cloud providers, that's not really a draw for us. It would actually be a net expense increase for us in the short term. So one of the things we're looking at and the reason we look at technologies like Cloud Foundry and some of the other emerging cloud technologies is it allows us to make our applications cloud ready so that you have the mobility for the applications. Historically, we never build applications that way. 12 factor didn't exist. And but even if we wanted to go to public cloud today, it doesn't make economic sense for us to do that for a lot of cases, even if you disallow some of the security issues. Then I think the other component, which is one we talked about and Bill helps comment this morning about public versus private and what does that mean for that to go away? What it means is some of those security issues, whether it's data locality issues, regulatory issues around data locality or even just how do you secure an application so that if you're running multi-tenant either in your own data center or in a public data center, once you've secured your application to the point where there you have complete isolation in a multi-tenant space, it doesn't matter where it's running, it could be running in your own data center, it could be running somewhere else and it's really just the economics that drive that. So I think when we look at multi-cloud, hybrid cloud, the strategy is in the short term around the economics and then longer term around a security posture that allows applications to be mobile. So some very good points. I would certainly disagree with the point that you should just stay on private. I understand visa might be very sensitive, but there are some workloads that probably belong in public, some workloads don't belong in public, right? So I would agree with that point, I guess. From point of view of building data centers, hell no. I don't think we wanna build another data center. It's very expensive. In terms of the operational expenses, I think cloud makes sense. Capax goes away, right? So I think that that's actually, yeah. I think, but maybe the idea of sort of this cloud burst and I'm gonna throw that buzzword there makes sense where if, like you said, if you're a cloud radio, then you can actually burst in the cloud and if you have some analytics jobs that require extra processing. But I can't imagine if you're a global organization not having a multi-cloud strategy. I have a CDN, we run one of the largest private networks in the world. And I can't imagine putting a data center somewhere in Asia or building a data center in Asia. I would rather use a cloud provider. Security-wise, I actually trust more Google, Amazon and Microsoft because I'm sure they will lose business if there is, you know, security issues. So I'm sure they thought of a lot of things that even on-prem we might not have thought of. All right, thanks, Audrey. I just have a question related to that. And I think you have something you wanna say, but I was talking to someone last year at a large enterprise that says they can't move to public cloud because they need the ability to sledgehammer their disk drives after they're done, for example, as opposed to, you know, making sure that whatever, even though they're securely erased or whatever, they need to be owned by the, you know, make sure I own- Shredded, I think. Shredded, yeah. There's actually industrial shredders that shred the disk drives. We do that. Does that apply? Does that apply in this, like, is that actually still a valid? Well, that was my point about security, right? Once you have the ability to run secure applications anywhere, it doesn't matter, right? You know, if the data's encrypted in a secure way. But security is not stagnant, right? I think the thing is that, and sorry, the point that I wanted to raise, it's interesting that actually the competition, Bloomberg and Thompson Reuters, they are saying the same thing, so that's a good thing, right? It's great, we are saying the same things. Because just 30 seconds at one point, another thing is that, and again, we were talking about it earlier, that in the center of all this, what we are talking about is big data, is all the different data that is coming in, and cloud is the enabler. Now, if you want to really utilize all the different unique sets, structured, unstructured, this, that, no matter what it is coming from, and really scale it the right way, cloud, public cloud is the way to go. You cannot just build all the security parameters in your data center. And yes, there would be 5,500 applications that you absolutely cannot move out of your boundaries, fine, but I think majority has to go that way. Is... Yeah, I think in general, I mean, perception is one thing, and in reality, I mean, the companies like Visa, it's not just one layer of customers. We have to ensure the trust all the way to downstream. So everybody understanding what is the security level of public cloud today, it's not there. Maybe after five years, it may be. That's the one thing. And the second thing is, in the Visa, we don't need to build the, I mean, data center. We already have a data center. We already have an infrastructure. We already invested infrastructure for next five years, maybe 10 years. So throwing out all that infrastructure and having talented people, IT talented people, and just going public cloud, I mean, there is no way. I mean, CAPEX and OPEX, when you compare, it doesn't add up, right? I mean, the CIO never agrees that. So there are multiple factors. It's not just like the technology that you're seeing there. You can go and consume it. But in the financial sense, when you go to your CIO and how OPEX and CAPEX works, and then what is the investment you already made, and what happens if you move, that didn't work out to us. So maybe if you're building a data center, I mean, from the scratch, maybe it's a good idea to go to public cloud. And once the security is no more concerned everybody understands that. Maybe at that time, we might go to public cloud. But today, where we are, where the industry is, what customer thinks about public cloud? So we are unable to move to public cloud. So that's... Yeah, but maybe the point on big data, which as Eve made, I think it's a very important point. So the data is new gold, right? That's what you always hear. And to process this data, you need a lot of technology. You need new technology. When we look at Tatooop, right now it's already kind of old. There are like 10 new frameworks which you can process your data even better. And you need an engine for each of your kind of type of data you have. Is it whatever form and whatever format fits best? So you need to build up all this technology in your own prime environment, whereas the cloud just allows you to scale. So I think technology, fostering leading edge technology for your data will give you competitive advantage by processing your data. And this is also some benefits which I think cloud gives you. And just another point that I want to add in there, again, in the support of public cloud is that community, right? So if you look at all the $100 billion industries that you have is all about community, how you build the community. And again, specifically in the financial services, if you really want to leverage and partner with the fintechs, integrate different things, you have to do that, right? So I totally get it. I mean, there are elements of data. I totally get it that you cannot repeat, cannot leave from your, for various reasons, yes. But again, we are talking about most of the datasets, most of the content that you have, how it will go and how it will transform. And I think the journey has to be, otherwise it's the cliche, it's being used a lot, but it is about disrupt or get disrupted is the narrative in there. So last year I was at a conference and a large company whose name starts with G had a little statement about that, basically private cloud is dead. And obviously they would, right? But their statement was that it's like the old power plants, 100 years ago, every company had a power plant sitting on their front line to power the building. And now that power is a commodity that can be centralized and same thing with, that was their argument there. From what you guys are saying, or some of what I'm hearing is that, today that's not possible because of existing legacy infrastructure and because of security concerns, but maybe in some short period of time in the future, maybe the private cloud will go the way of the on-premise power generator, you know what I'm saying? Yeah, and I think the interesting thing that Asif brought up about data, there's a lot of different kinds of data in the financial industry. So like our data is predominantly not our data, it's client data. We actually have to sign agreements with our clients to use the data even to process on their behalf. And most of those contracts were written before big data, the internet, et cetera, all that kind of stuff. So we literally have to rewrite contracts with our clients to be able to do some of those kinds of analytics. I think also, like in the case of Bloomberg and Thompson, they own the data, that's their stock and trade. We were talking a little bit about before the series, to really get to the next level for all these different kinds of data, you have to have some way to secure at the unit level that data, like maybe water market or have some other kind of way to transact on that data. So if a client wants to lease their rights, they can do that. If you want to know who's using your data and get a piece of that every time it moves, you can do that. And that infrastructure just doesn't exist today. So one of the other reasons we keep our systems in house is because of the way we have to secure data and we don't have any way to control it when it gets into the wild, wild internet. Well, one question about that, because we see that often, especially in Europe. What is the request of having this data private? Is it because the technology is missing, which can kind of prove that it's secure enough? Or is it because of the gut feeling? Some of it is perception, for sure. And you have to have a business reason to go through the effort to convince your internal and external compliance and regulators that it's a great idea and it requires a lot of convincing. So there's got to be a real business reason to do it otherwise, what's the point, right? But if you had a way to secure that data and track it and manage it so that you know how it's being used, if it has a financial value, you extract that value when it's appropriate. That's a different story. And I think when that comes along, I think you're probably going to see more pressure to put some of this data in places that could be more readily accessible. Yeah, one point to add is I agree. It's a perception is the key thing, actually, today. Security, I mean, so you can ensure security in a public cloud as much as in a private cloud. But the perception, the perception is not there yet, right? That's a key thing. So just to add another point on the private cloud. So we have to see, I mean, some of the early adopters, like Netflix and Tommy and few other companies, they are moving back to private cloud. So we have to study why they are doing that. So why early adopters? Some of the early adopters, they were in the public cloud for a very long time and why they're planning to build their private cloud. That's actually, recently, I was in a CIO conference, I was talking to a customer, they're planning to build from the scratch using the latest technology. So, I mean, it's a perception and, I mean. Yeah, well, one of them is data gravity. Once you build up a significant data set, the meter is running forever. Exactly. You know, the only way to get it back, you know, like even now, Amazon's got things like snowball and stuff like that, but even that's just like pilot. How do you get your data back? You know, and the meter keeps running, but I think also you need to profile your applications from a financial perspective. There's only a certain percentage that makes sense economically. And, you know, in particular, some of the larger cloud providers have made their pricing so complex and opaque that you literally can't figure out how much it's gonna cost you. So unless you're big enough and you're doing enough business to write a bespoke contract that locks in a certain rate, you know, which Netflix did and some of the other providers, you know, it could very well be more expensive. And that's probably the objective is to get that. That's a great point, right? Because I think, so if you look at the challenges, I feel that you're spot on that, I feel that most of the, of course, Microsoft knows B2B really well, that they have been doing it for many years. But if you look at all the other, most of the new vendors who are big, i.e., the AWS or the GCP, which is coming up a lot, Google Cloud, they rock when it comes to B2C models, to selling it to you and me as individuals. But when it comes to B2B model, they're getting there, they are, you know, and again, there is a challenge of what is the right business model to actually do that. I think what they have as tools is incredible. I mean, even from a GCP perspective, TensorFlow and some of the capabilities, Kubernetes we were talking about earlier, of how you can actually, that can support multi-cloud is fascinating. But back to your point, David, I think business model of how you actually take it and make it and put it into a production is an issue that we are all facing and trying to solve. But, I mean, as we are in the Cloud Foundry Summit. Yeah. Aren't technologies and what the standardization which Cloud Foundry offers you, is this not a bridge to public and private and which allows us then also to switch the providers, switch the models, whatever applies right now, also maybe based on where your company stands. Right, so that was... With the culture, yeah. So, do we have time or no? Yeah, we have time. There's one more point to consider which is disaster recovery. Which is, right now, in my case, I don't have presence in certain geographical regions. And for me, even if it's just for DR purposes, Cloud makes a lot of sense because then I can have that presence there and that locality. So, you know, and I can imagine building a data center for that. Right? If we're talking about economics, you know, building a DR in different regions, I think that's pretty attractive. You know, even if it just comes up on demand, I think that use case is actually quite good. And if I use Cloud Foundry, it will, you know, magically appear, right? So, you know, you can just push up and that's it. Very good. So, while we're on the... You've got your 7,000 line YAM already. Yes. There you go. So then another... If you alluded to it already, the ability to move from Cloud to Cloud is one way to ensure avoiding vendor lock-in. And how important... I'm very curious to know how important vendor lock-in or avoidance of vendor lock-in is for you folks. Is that it? I can already see this with AWS, with all the wonderful features and candies that AWS has. You know, the vendor lock-in is happening right now, so it's definitely a big issue and certainly things like Cloud Foundry or Kubernetes or, you know, other platforms abstract in a way. It makes perfect sense. You know, I don't think we have the full parity of all the services, but at the same time, if you don't think of it, then, you know, you're basically going to be locked in and the price in and everything else is going to, you know, get you down. I think that's why Cloud Foundry or even containers, the new containers that are coming out, the Kubernetes and all, whatever it is that actually keeps you agnostic is priority number one. And if any vendor, any cloud provider is thinking any other way, they will, I don't think they can survive long-term because it is about being agnostic. It is about being actually, again, building that community of cloud and that's the only way forward, in my opinion. Yeah, I mean, for us, actually, vendor lock-in is not at all allowed. I mean, so we don't want vendor to dictate us. So we want to build a solution based on the technologies that we like, the technologies which suits to us. So in fact, we are building actually multiple clouds. I mean, so the cloud with a, I mean, VM orchestration, the cloud with the Cloud Foundry, the cloud with the Docker and Swarm. So we are building multiple clouds based on requirements from different product teams. And same times, we are trying to make sure we are not locked into any technology that's particularly going to dictate us. Yeah, I think, I mean, I can just underline what they said. And I think one further point which also the foundation brings into place is a area where people like you talk together and form the new requirements to the product. And I think that's, that uniquely shows that the vendor lock-in is pretty, pretty, yeah, not here. Yeah, I think even, you know, this ties in with the comment this morning about, you know, not private or public, but hybrid is, you know, certainly for us, you know, we've got the same concern you do. We've got capital investments. But even once you start, you know, rolling off some of those assets, you want mobility, right? If I want to be able to say, okay, I can put it in this public or in this externally hosted area, or I may want to put it in my own data center. From my application team's perspective, it shouldn't matter, right? There shouldn't be a lift. There shouldn't be an issue. There's probably going to be some work behind the scenes in terms from a compliance perspective. But, you know, that's really the goal is, whether it's moving from one external cloud provider to another or to my own space or wherever, I want that mobility. And I'm not going to get that. Even, you know, using something like Cloud Foundry, because it's not just, you know, Amazon versus Azure versus Google. It's also, can I run it on my own infrastructure? None of those guys offer an internal, in any true sense, something that I'd run in my own data center. Very good. Okay, so we actually have... We have five minutes left, and I'd like to at least invite if there's an audience question or two, if anybody has some burning question. And then please step up to the mic and we'll cover it. Thanks, guys. So you have NASDAQ has something that runs in AWS, QCloud, Cap Ones on stage at AWS. They're a Visa, you know, brand. So just curious what you're thinking is there. What's the specific question? I just wanted to... I think he's countering some of the comments we're making here that public cloud is a little farther off. Ah, okay. No, definitely. I think the moment... So my view is that everybody is definitely going towards the public. There is no question about it. I mean, I'm sure that all of us are in some way, shape or form. How far are we in the journey, right? We also... Thomson Reuters also just not from a financial services, but in some other areas, like we have other portfolios like tax and accounting and others, maybe we are further down the road on the cloud story than we are on the financial, just because of the complexity of where it is, does not mean that we are not on the cloud on the financial side. And to your point, there are a lot of other, you know, big names who are already using cloud. So I think it goes back to, again, the complexity of entitlements of, you know, 400 plus exchanges, how you actually... What is... What can I entitle you versus what I can entitle the rest of the room and so on forth and can public cloud handle that? What are the issues in there? Because, again, we also, and I'm using our good friend Bloomberg and Thomson, we have a lot of different data sets, a lot of different contents from 400 plus exchanges. I'm sure Bloomberg has the same thing. And everyone has different entitlement things. So how to handle that? Those are also... Yeah, and one thing I would say, I've heard some about the cap one story, you know, just in terms of what's been spoken about publicly, there's certainly, you know, in the spectrum, they're, you know, an outlier in terms of being very aggressive about converting their internal workforce to a completely agile workforce. They've got a pretty interesting story. You know, I think the folks from all state as well, you know, I don't know how many caught their presentation in the afternoon about their culture, but that's the same kind of thing that cap one is doing. And I think, you know, I would be, you know, I'm trying to remember if I heard them say specifically, there's some things that are never going to run in the cloud. I think they're going to say the same thing that we're saying, but they're definitely way farther out in terms of what they're willing to put on public cloud now. And they've done, you know, from an economics perspective, that's the conversation we were saying. They did the math. They got their compliance and legal and regulatory and all those folks lined up. That's a business strategy for them. I don't know how it works out, but, you know, I think you can put all the pieces together. I don't think anybody here said, no, you can't do it. But, you know, it's just, it's not part of our individual business strategies. Very good. So we have time for one more question briefly. So go ahead. Great. Thanks, gentlemen. And just picking up a bit on that theme, I think there was the reference to Google saying it's all about public. And Amazon used to really hard message if it's all about public. But given you're all represented in the financial industry, all information's not really valued and were created equally. And I was just wondering, you know, just like today, there are still industries that generate their own power, that have substantial backup power infrastructure because their industry requires that. And I was just curious if you felt looking forward that that was likely to be an element that public cloud would be relevant to different industries in different ways. Well, yeah, absolutely. I think it's going to be the relevancy is going to be different. One of the interesting things, you know, and again, I mentioned, you know, BNY Mellon is really a back office provider. We do a lot of different kinds of financial products, but one of our core businesses is doing business for other financial services companies. So we don't own the data. One of the interesting things that you're going to see is with blockchain and other types of peer-to-peer technologies, you know, where the ownership of the data lies, how it gets moved from one place to another, that's going to be, that's more likely to be a disruptive force in the financial industry than things like, you know, startups, you know, launching a new financial services company on Amazon and competing with one of us. That's not going to happen for a while. They're regulatory and other reasons why that's difficult to do, not impossible, but difficult. But I think some of the other technologies like blockchain, I think have more potential to do things like that. Yeah, the way I see is, I mean, in next few years, one or the other way, every enterprise try to move into the cloud. Whether I can't say that it's a private cloud as public cloud, but the cloud is the essence. So actually before coming to Visa, I was with Oracle and I was building the public cloud. So the security point of view, I don't see much difference. It's mainly perception. So the data, customer data that is in the public cloud may be completely encrypted. The encryption key will be with the customer. So vendor cannot do anything. So your data, you own it. You have the key. So the security point of view, I don't see much difference. But I mean, as we said, regulatory point of view, so there may be applications that may need to stay in private and rest of the application might move into public. I would say that very quickly it's the BlackBerry model that we have talked about previously also. And what I say is that living in financial services for many years, five, six years ago, BlackBerry used to be the gold. Locked in, you cannot do anything. Everything was locked in. Now five years later, everybody has a smartphone and every that perception has gone away. Same thing will happen with public cloud, in my opinion, with the right time. I don't know if it's five years, two years, three years, but it is the BlackBerry model. And people who don't go with the smartphone model will face, I feel, the same thing as BlackBerry model. What happened to BlackBerry? Very good. Okay, well our time is up for this afternoon, so I'd like to thank the panelists who showed up and participated. And thanks everyone for being here. And I guess next up is the keynote, so thank you.