 Live from Santa Clara, California, it's theCUBE covering Open Networking Summit 2017, brought to you by the Linux Foundation. Hey, welcome back everybody. Jeff Frick here with theCUBE. We're in Santa Clara, California at the Open Networking Summit 2017. Really happy to be joined by my co-host for the next couple of days, Scott Rainovich. And we've been talking a lot of providers and technical people, but now we want to talk to customers, right? We love talking to customers and we're really excited to have Roshesh Jethi. He's the SVP of head and R&D for the Americas for Amadeus, which is a big travel company. Welcome. Thank you so much, Jeff. Thank you, Scott. So like I said, we love to talk to a practitioner. So you're out on the front lines. You're seeing all this kind of talk of software defined and software defined networking. From your point of view, how real is it? Where are we on this journey? What do you see from your point of view? Super real. Have you searched for a flight lately? I have searched for a flight. Excellent. Then I'm proud to tell you that your flight search very likely was powered by Amadeus and it's running on a software defined data center. Completely. So this stuff is real. We are, I believe one of the first companies who have actually taken this from what was a very strong academic kind of research project onto like the startup ecosystem. But we're actually out there deploying it, running real world business, using a very purposeful and deliberate software defined strategy. And it's interesting because you said before we got on camera that you guys are actually, you know, very active participants in the open source movement and development of this stuff. You're not just kind of a participant waiting on the wings for this stuff to get developed. I mean, absolutely. And that's to me, that's one of the ways in which, you know, if you're serious about open source, you have to use it. You can't just talk about it. You can't just, you know, say it looks like a nice idea and you have to get out there and get your hands dirty and do it. But the other thing also is you have to contribute back. I think that's a big tenet of the open source community. And, you know, we all, and certainly the company we grew up and we've seen tech evolve through the ages. And a big part, especially in the last, you know, 10 years or so has been this open source movement and it's contributing back. It's one of the reasons I'm here. It's one of the reasons the conference organizers invited me to actually talk about how we use open source and software-defined strategy for our technology. That's cool. So where do you run the software? You run it private cloud, public cloud? You guys build your own data centers? So how do you run it? Quick history lesson or quick history, you know? Yeah, we should bring it back up. What kind of the, first off, where is Alameda today? For people that aren't familiar with the company. Right, so we're actually a 30-year-old company. We're celebrating our 30th birthday this year. The company was started in the late 80s as a consortium between four leading European airlines, Lufthansa, Air France, Iberia and Scandinavia. And over the years, you know, so we started off which was very typical at that time as a mainframe shop. And that's where a lot of our core systems were built. We were a big provider of technology in general to the travel industry even though we were founded by airlines. So to put it in perspective, we carry about 95% of the world's scheduled commercial seats, airline seats on our platform. 95%. 95%. So we work with... Are available to purchase. You know, obviously 95% of the purchases don't go through yours as well. Correct, they are available. We are used by over 90,000 travel agents, retail travel agents, corporate travel, online travel. And we work with over, like I said, 700 airlines to offer the inventory. So chances are, you know, that if you travel on an airplane, you know, very good chances that our software was used to make the reservation. We also have airline IT systems and hotel IT systems. And we work with airports. And this is where we do departure control, flight management, baggage reconsolation, a lot of the backend processes. And we started the company essentially runs, we write our own software. We offer it as a service from day one. So we are one of the oldest software as a service providers in the industry. And obviously when we got started to do that, you had to own your own infrastructure. So we are pretty good at it. We have very strong kind of technical chops. We have a large data center outside Munich Airport and a bunch of smaller data centers all over the world. And what we're doing now is really, you know, very deliberately making the journey towards a cloud, both our private cloud. So taking our own infrastructure, virtualizing it and making it available as a service for our own applications. And then where it makes sense to leverage, you know, public cloud infrastructures that are available. So different apps and different clouds, is that? Different apps, different clouds based on customer preferences. The core reservation booking engines, they are on-prem in our own private cloud because we do have a lot of regulatory, security, privacy, and our considerations. So that stuff we keep kind of closed where we can keep a very watchful eye on it. But there are a lot of transactions. We're also talking about, you know, the volume of searches has grown up, right? Obviously Google's seen a lot of search volume. If you look at our business, it used to be when you wanted to book a flight, you'd go to a travel agent and they would look at a bunch of flight options and you'd pick one. And about 20 years ago, it was, we called it the look-to-book ratio. So you'd look at 10 to 20 options and you'd book one. You want to guess what it is today? The look-to-book ratio was 20 to one. 20 to one. That's got to be way higher. It's got to be 80 to one. It's more like 1,000 to one. 1,000 to one. And, you know, it's partly people like you and I who have a spare moment and have a vacation in mind and we are looking at options. But keep in mind, anytime you search, it has to come into our systems. We have to configure the journey. We have to price it. We have to make sure it's available before we offer it up here, right? So it's very transaction-compute-intensive. Even before it touches any of the back-ends where we do core kind of booking and passenger processing. And so to handle that scale, you know, those are the kind of very logical applications that make sense for the public cloud. And those are the ones that we've looked to move. Certainly for customers, we are a global company. We have customers all over the world. Some customers want to have some of these systems closer to their geographic locations. We look at all use cases kind of. So that's amazing to think of. You know, these things have so changed behavior in the way that we interact. I assume that 20 to one was a function because you would sit down. You know, you sit down at your desk. Time to book that flight. And you know, maybe you don't get it done that day. You come back two or three times. But as you said, now it's grabbing little bits of time throughout the day whenever we can. But do you get paid like on some regular subscription or do you get paid on the transaction? Does that just increase your overhead? Whatever the ratio, 22,000 is? Absolutely, no. Our business model's been very consistent from day one. We get paid on the number of bookings we make and the number of people at board aircraft. I mean, roughly speaking, we have smaller lines of businesses but those are our two main revenue drivers. So if you see a lot of transaction volume up front, but it doesn't translate to a booking, which logically it won't. Yes, that's no extra revenue for us but we still have to service that volume because that's eventually the fund has gotten wider. Right, right. And so it makes sense to do that in the most cost-efficient manner. But without compromising quality, without compromising speed, because I mean, if you're like me, if you have to wait for more than two or three seconds, you're like, ah, I'm moving on. Oh, two seconds, it's milliseconds, isn't it? Absolutely. And by the way, I still don't always find the flight I want. So where are those extra flights? Can you provide those for service? Yeah, that extra 5%, those are under. No, that's a very different say. It's got nothing to do with open source and kind of what we're talking about here, but there's a lot of plot we're doing in there from an engineering perspective, which is looking at, for example, machine learning algorithms, and what you said is actually a very common complaint, is how do I find kind of the right set of flights? And more importantly, if you have certain preferences, you know, with airports or airlines or loyalty programs or time of day, you know, how do I provide you kind of context-sensitive results, right? And we are doing a lot of kind of core R&D work for that, but our customers are doing amazing work as well. You know, Kayak is one of our customers very close to kind of our offices in the Boston area, right? And they do pretty amazing work in terms of getting the context right and then applying machine learning technologies and artificial intelligence. It's very, very early days, but very exciting, very promising. Yeah, one of the cool features I like are these like fare alerts. I don't know if you use them, it tells you predicts, you know, this is going to go up, you better book now, or it's, you know, wait. Do you guys do that sort of thing too? Our customers do that. So, you know, we have a very simple model. Our customers are travel agencies, online, the American Express is, you know, Expedia's, the meta-searchers like Kayak, SkyScan, et cetera. The airlines themselves, whose product we host in our system and we sell. So a lot of our engineering work is done into offering kind of core innovation so that they can offer products for people like you and me, their customers, you know, that are the best products out there. So we focus on enabling them and then, at an operational level, we like, you know, try to do it in the most efficient manner and the most, you know, future proof that we can think of. What about, sorry, what about security? I mean, it sounds like a lot of sensitive data changing hands here, right? Where people are going to sit on an airplane, where they're going, but you must have incredible security demands on your data now. Yes. I mean, you understand it's one of the areas, obviously it's paramount to us and I mean, the good news is, look, we've been in this business for 30 years. We have really deep domain expertise in that. And also you'll understand why I don't, wouldn't want to talk too much about what we do and how we do it, but absolutely that's one of the- You just lock it down. The prime drivers of, you know, everything we think, you know, all the way from application design to things like, you know, the infrastructure planning and design to the physical level. I mean, everything you can think of and probably a couple of things you may not think of. Hopefully, hopefully a few things we didn't think of. So where do you go next? It sounds like, you know, you're enabling a lot of innovation on your partner side. You just mentioned, you know, Kayak and people writing some of the machine learning and AI algorithms to help the end traveler, you know, find what they're looking for. What are you guys concentrating? You've been at it for 30 years. What are some of the next big hurdles that you're looking to take down? You know, it's wonderful. I think being close to our customers and one of the reasons I'm in Boston, we're a European company. We're actually headquartered in Madrid. Our core engineering team, our central engineering team is in France. The reason I'm in Boston and my team's in Boston is, you know, we started doing a lot of business here in North America and we try to stay very close to our customers and when you listen carefully, and that's why we have two ears in one mouth, which is, you know, hopefully try to listen a lot. You do see their pain points. You do see where they're going with kind of their business and it gives us a chance to have a front row seat in designing new products that they can use. So to me, it's kind of two prong. One is, you want to offer the best technology we can to our customers at the best price point we can. You know, and obviously by now you've figured out it's mission critical stuff has to always be on. So keeping those kind of boundary conditions in mind, you want to be the best technology provider and then we want to innovate. So one of the things I'm seeing at this conference, there's a lot of friends from the service providers. We're talking about 5G technology and so with connected cars, with virtual reality, I mean, these are all trends that are going to impact us as travelers in a positive way. And so we have a dedicated kind of innovation team across all our business lines. We do a lot of work with kind of academic institutions with ETH and Zurek, with MIT here, close to my office in Boston. And there's just a talk full of possibilities in terms of what can be done. All right, I'll give you the last word. Impressions on the show. You know, what do you get out of a show like this? Why is it important for you to come? Oh, it's amazing. I mean, this morning, Martin Masato was there and he's the one, he's called kind of the granddaddy of software defined networking and virtualization. He's not that old yet, but he's going to like seeing that. Well, I read that at the Guardian, so it was one of the newspapers. But the fact is, we use NSX for virtualization now entire data center. We have close to 20,000 infrastructure devices. All our computers virtualize, 100% of it, right? And it's all using NSX from VMware, right? Now, this was a brilliant idea by a extremely intelligent and persuasive graduate to understand for, you know, 15 years ago that is, as he announced this morning, he's a billion-dollar business today, right? And we are actually using that technology and it's very real to process all of this. So it's great to be able to see what people like him, Amin from Google, is a great partner of ours. We use Kubernetes for kind of the container deployment strategy for our cloud network. You hear him speak about what they're thinking about in terms of investments and how the network is going to essentially drive the movement of data. I mean, it's just a phenomenon to get that thought leadership. I'm obviously very honored and privileged to be presenting to this audience and to share our thoughts and what we are doing. So, and just to see a lot of the buzz around here and what wonderful ideas are happening in the valley. This is where so much action is always going on. Great, great, great summary. Well, I'm glad you could take a few minutes and stop by theCUBE. It's completely my pleasure. Thank you very much, great meeting you. And have a great festival. All right, he's Roshesh. He's Scott, I'm Jeff, you're watching theCUBE from Open Networking Summit 2017 in Santa Clara. We'll be back after this short break. Thanks for watching. Help to bring a product called Checkpoint to Canada, firewalls, URL filtering, that kind of stuff. But you're also.