 Live from Bahrain, it's theCUBE, covering AWS Summit Bahrain. Brought to you by Amazon Web Services. Okay, welcome back everyone. We are here live in Bahrain for the exclusive CUBE coverage of AWS Summit here in the region. Obviously huge news, Amazon's having a region here, a full region. That's going to create a lot of connections, new opportunities, and hopefully make the life easier for all the developers and whatnot. Great guest here. So we're just talking over the game on camera about all the exciting developments on Amazon. We've got Abdul Rahman, who's the group EVP of tech at the ATG, which is the Al Atigachar Travel Group in Saudi Arabia. Yep. Thanks for joining me today. Thanks a lot. So I'll quickly fast forward. You guys started in 2015, programming in the cloud. You're like, we were late. I think that's actually a good time because Amazon had a lot of mature services ready. Went from zero to billions in revenue. Correct. Really a big success story. That's a large scale, all cloud-based, right? Yep, correct. Tell your story. What do you guys do real quick? Take a minute to explain your group, what you guys do, and then what were the architectural things you decided? How did you get to growth? So we are a 40 years old company. We started in 1979. We are the largest travel and tourism company in the Middle East. We went public through our IPO in 2012, and in 2015, our new board and new management, including myself, we started building our 10 year strategic plan. And we said we need to diversify our investment. So it mandated that we need to have an online presence. In 2015, we had a choice to build our online presence, which is very late, either on-premise using building a data center or we go to the cloud. We had multiple metrics, including the cost, efficiency, including scalability, security, and so on. And all these metrics, when we compared on-premise versus cloud, cloud always win. And we selected Amazon to build our online presence. And beginning of 2015, we had zero presence, zero revenue. Our total revenue from the classic legacy systems for the retail was almost $2 billion, but we have zero revenue from the online. We were able within six weeks to build the growth of concept and launch it immediately. And we started heavily investing in various components from back-end, front-end, DevOps, and so on. And this year we anticipate we're going to be generating more than $2 billion of revenue. That's about $450 billion. Online. Online only. Via cloud. Exactly. Only on Amazon. And for us, that has been the best success story we had for years. It's an amazing success story, actually. We look backward to our decision back then. I'll pray for you. I'm actually really amazing. This is something that I think people don't really understand what about the cloud and certainly Amazon and the kind of scale that you can get if you get something right, both on the business model side and architecturally. You could be a unicorn. You're really a unicorn in revenue. That's the word that they hear in startup world. Unicorn, but mostly that's stock value. It's not actually real cash. And how many years? This is pretty phenomenal. This is the entrepreneurial dream that is now a reality. This is the story here. Exactly. I'm happy you mentioned that. We actually, when we started this venture, we said to the founders, you guys are a startup. We rented out in 2015 a garage, literally. A warehouse. Get out of the way. A very old warehouse. We brought like five guys. You are the core team. We told them you are a startup. Give us whatever you want to do. And it has been very successful since then. It's kind of like the Steve Jobs story. You got Apple with the Mac 2, and then the little group over here doing the Macintosh. That's your group. Because you got to get out of their way. It's a mindset. I want to ask you that. It was one of my questions, but we got there a little early. But this is a cultural shift. Cloud is a different mindset. It's not the old way of planning, team building. It really is a different dynamic, both execution-wise, but team makeup. Can you share that piece of it? We gave our founders complete freedom and how they're going to make up their management style. So we have a complete agile team. We have diverse geographical locations. We have people from India, developers in Egypt, in Dubai, in Saudi, and they all work and collaborate using DevOps tools from Amazon. So we divide the workload, our product teams, weekly feature list, they tell us when they want to launch every two weeks or three weeks a new version of the website or the mobile apps. So we have a completely agile development methodology and we give our new venture a truly startup culture. And the key for you, if I get this right, is to have executive leadership say, we're doing this. Was that in place? Did you drive that? Absolutely. So when our board said, tell us the new board in 2015, guys, we don't have an online, go and get it. Me and the CEO said the best way to do it is just spin off a completely different unit, completely independent, start a mentality into renewals and told them, guys, sky's limit. We need to be the number one player in the Middle East. So I got to dig deeper because I love, it's all sexy and great story when you say, this is how we started and we finished strong. But as Andy Jassy would say, the CEO of AWS, the learnings in the middle, the ups and downs, as you figure things out, is because a lot of things about cloud's iteration, because you have the ability to move very fast and you get smart people together. So there's a glorious start and a glorious outcome, but in the middle is the experimentation. That's where the real work gets done. Can you share some of the learnings? Was it a technology selection? Did you really, did you have more queuing, more database? As you start to play with Amazon, this becomes actually a business process. Oh, the biggest? Playing with the different pieces and what services are right for which process? Can you share something? Correct. So our biggest challenge was finding the right skillset for people who understand how Amazon AWS work. In the Middle East, we don't have that many skillsets or skillful people. So we had to wait, train the people, send them to Amazon workshops, be very patient with the mistakes. We don't mind people refactoring all the old code every month we start from scratch. We were very aware that this is what we are doing. It's never been done before in the Middle East. And what we have developed in terms of, for example, the big data, the big data platform we built today is one of the largest. We are processing terabytes of data every week. It's one of the largest in the Middle East. The number of developers we have today, more than 500 working on AWS. I don't think any company in the Middle East have that number of developers working on this platform. So we're very proud that we gave our developers the trust and we are aware that you need to fail fast. Learn and quickly adapt. And it's a contagious mindset too when you start seeing success. So talk about some of the architectural, talk about the stack that you're using. Obviously you must be using a variety of the Amazon goodness, EC2S3 obvious. Are you guys using the queuing, are you using Kinesis? Can you talk about some of the architectural things if you can? So the front end that we have today is completely built on Node.js and Angular.js. So it's very fast, very agile. Our back end is built on Java. Most of the code built on Java. We have multiple messaging buses that asynchronous mode. So whenever there is something that needs to be given to a certain component, we don't have to wait for serial queuing, it's all parallel. At the same time, we have a lot of auto scaling components. One of the examples I gave earlier today is that we had the beginning of this summer, we had so many marketing campaigns and we were surprised by how successful these marketing campaigns. We have noticed in one marketing campaign that our demand from our customer have reached 300% within 24 hours. And the auto scaling that we have in place have been very successful. We were able to immediately meet that demand. Talk about how good the auto scaling is. Isn't that a relief? Absolutely. I mean, explain how it works because essentially when the demand comes in, explain how it works. So just to give an example, if we had this infrastructure on premise, we would have needed six weeks to procure a new infrastructure, install it, configure it, and we would have lost all these six weeks of revenue. And then by the way, you would have lost the first 24 hour surge, then you'd go over build and then wait around and then not know if you over provision. Absolutely. This is the old way. The new way is you configure auto scaling based on policy, and then it just spins up resources while you're sleeping. Exactly. So in a few seconds, the auto scaling fires up a lot of instances and we immediately cope with the demand. You know, it's a funny, you mentioned that. One of the comments we have inside our company is, you know, you're successful online when you're making money while you're sleeping. And you know, and if you have auto scaling and things of that nature, these things are programmatic. This is what Elastic is all about. This is what coders, not system administrators do. And once they do it, they're highly motivated not to manage it again. Correct, absolutely. Again, this is back to the culture of DevOps. Yep, yep. How have you guys innovated on that piece? Can you give some other examples? Yes, so today we have, our big data has feeds from all the vibes from the big social networks, Twitter and Facebook and also from Google. And we have all this analytical data into our big data and we analyze all our customer behavior, what they're looking for, what kind of destinations, holidays, business travel. And we try to adapt every two, three weeks our product and services to meet our customer demand. Next year, we're going to be launching our machine learning and AI infrastructure. This way we will be able to do real-time, predictive analysis, and we will be able to serve each customer unique, fully personalized, customized, web-based and experienced. We will be able to exceed our customer expectations and we'll be able to give what our customer, exactly what they're looking for. Abdul, I got to ask you a personal question. Sure. What are you most proud of of this success story? What are some of the things that you look back and say, wow, we really knocked it out of the park, we did great on this. And then an example where you had a good learning experience, maybe a trip and a fall, that was a learning opportunity. What are you most proud of and areas that you learned the most about from tripping and falling and failure? Yep. I think the most I can brought up is we have gathered great minds and we have created great culture. I think great companies have great people behind them. And this is how I've learned from reading the stories of Apple or Microsoft or Google and so on. So I think we've been very successful in this area in the Middle East where the resources are very scarce and the ability to attract very smart people is very difficult to bring them in the Middle East. And I think we've been very successful in that regard. We've been able to gather a lot of smart people and create great culture. You know, Mark Andreessen wrote that article, book about, or maybe it was a tweet, I can't remember, the 10X engineer. Yep. And that concept is one engineer that does cloud and DevOps right is worth 10 engineers in the world. And so if you can collect a selection of these 10X multipliers that can do architecture. Right. Now I personally believe that the full stack developer is might be obsolete with the cloud or reduce the requirement for full stack developer, but you'll still need full stack developers for cloud in general, but you don't need to stockpile full stack developers. True, true. I agree. If you have good full stack developers, you they can hire application developers. True. Because the full stack takes care of all the scale. Exactly. You can always repurpose those guys and upskill them to do something different. Instead of being a full stack, you really want to focus on solution developer. Google's proven this with their SRE, if you've seen they have operators and developers. And this as you scale, you're operating infrastructure or you're writing code for applications. All right, so what's the learnings that have been magnified for you? In the middle of the journey here is always the situation where you had to take care of personnel issue or technology selection tweak or change, iteration, I won't say pivot because people don't pivot when they're succeeding. It's just navigating through the journey. What was something that you've experienced that was magnified in the learnings that have helped you get better? Yeah. I believe that the multicultural and the multinationalities and multidiscipline and people coming from different backgrounds. We have people from Asia, from Europe, from the US in our company. And this helped having different backgrounds, different experiences. And this has helped us to build a nice multidimensional solutions. And people have been able to share this experience in a very nice way. That's great. Abdul, thanks so much for sharing. Taking the time here on theCUBE and sharing your insight and amazing success story. Congratulations to you and your team. Really love to hear these amazing success stories, essentially building from zero start online to billions in revenue. That's an amazing success story. Thank you very much for having me. It certainly is great. Exclusive coverage here. We are in Bahrain, this exclusive CUBE coverage. I'm John Furrier. You can reach me on Twitter at Furrier or just search my name, reach out to me. Let me know what you think. Stay with us for more coverage after this break.