 Live from New York City, it's The Cube at Big Data NYC 2014. Brought to you by headline sponsor, Juan Disco, with support from EMC, Mark Logic and TerraData, with hosts, Dave Vellante and Jeff Kelly. Welcome back to New York City, everybody. This is The Cube, we're at Big Data NYC, concurrent with Hadoop World and Strata. Manny Chabras here, he's the CEO of Cloudwick, Cube alone, Manny, great to see you again. Same here, Dave. I mean, see you and I've been sort of the old warriors here, like five years for you guys and five years for us coming to Strata. It's awesome, it was fun seeing you last night. We had a great conversation and our Big Data Capital Markets event and a big party afterwards, so thanks very much for coming and I hope you had a good time. Thanks a lot for inviting us. Yeah, so we heard from Jeff Kelly yesterday that a lot of the dough being made in Big Data is in services. Big time. About 45% of it. And I don't see that changing any time soon, do you? I think Dave, this is the first time in the technology history where the services is leading to technology. Earlier it was hardware and software and the services used to be the third piece which used to come after, but this is the first time in the history of technology where services becomes first and then you follow with hardware and software piece. Yeah, it's kind of was, we went from break fix to implementation and now it's like, what do I do? I call guys like you, so take us through. You get a call from a customer. Give us a typical scenario. We typically don't come with the customer directly. We are very aligned with the partner channel. We are being brought by Cloudera, Hotlineworks, DataStacks of the world because what they can do very well is that they can take the customer from the use case discovery to the pre-production. And then we come in the play is on the production side of the business because scaling from 40 nodes to 400 nodes is not an easy enterprise and our focus is enterprise fortune 1,000. That is where we think the really, really need is of help on the DevOps side of the engineering side of the business and that has been our focus. So talk a little bit more about the DevOps side. So you've got an affinity with the DevOps folks to talk about the state of DevOps in that Fortune 1,000. This is a very interesting day. I mean, we had a great discussion last night on this thing. One of the things which has happened in the last 20 or years ever since we moved from relational databases is the outsourcing piece. The outsourcing has totally hollered the enterprise IT. There is no skill set available in the enterprise IT today to manage the databases. And so the DevOps piece which has to basically do all the builds every 24 hour, 24 by seven time is very difficult for the enterprise because not having the skill set. Whereas Web 2.0 companies are basically based upon that model. And the enterprise is very, very having hard challenge identifying how to do those things. And DevOps is very required from these things because if you're a bank, you have to kind of move. You have to basically scan checks. You have to basically provide all the services which come in this digital world. And those are all relying on the latest builds every night. So that is a big, big chasm out there. And so is it a mindset issue? Or is it simply is it also again that just don't have the skill set to be able to do that? Or is it also kind of the change management challenge? It's always difficult to change if you've been doing things a certain way. Specifically, I mean, banking obviously a very mature industry. I think everything, it's the thing in the enterprise this is the one of the biggest disruption. I think the people transformation, the process transformation is entirely new. It's extremely, extremely complex for them because they have been doing a line of relation databases and preset worlds. But now you have to change according to everything. I mean, you know, and we know the thing and MapReduce used to be the big darling, you know, last strata. Now the spark. And then also there is no place to go anywhere. You can learn Spark, you can call somebody. You know, so it is very difficult. Like a Hadoop Spark is where Cloud era, Hot and Works, and then DataSecSparks, Cassandra, and so I mean, the three different entities supporting Spark. So how do you do those? Those kind of challenges in the enterprise have been extremely difficult. You know, earlier there's one call to make to Oracle to solve all the problems, but that's not what's happening. Right. So a lot of dissonance in the world. So you got IT folks, you got sort of the old and the new, and you know, the Hadoop guys saying, hey, this is how you got to do it. And the traditional IT guys go, whoa, whoa, whoa, wait a minute, you know, especially in the Fortune 1000, that's not what we do around here. We have to make sure that it's, you know, available, reliable, recoverable. Meanwhile, I get the business guys that just, you know, they want the results. So there's a dissonance between, there's always been between IT and business. And then within IT, how are you seeing that? And how do you help resolve those? You exactly point, that's one of the biggest things. The line of business guys have intuitive understanding of the markets now, because they see these patterns emerging out of Silicon Valley, where their own core business is basically being taken by these upstarts. And these upstarts are data driven business. They capture data at every transaction of the way. Now, if you enterprise your line of business, you know that is what's going to happen. But the IT is a tough one to crack. So one of the things which basically we come and do is that we have developed this tool where we capture all this base information. And our DevOps folks is called Cloudric Big Loop. And basically is an open platform for sharing the operational intelligence of the system with the vendors, like the Cloudira, Hortonworks, DataStacks, and the customer. And what we try to do at the first level is to establish the baseline, what your clusters are, what configuration, what hardware, what networking, you know, to get all the operational baseline. Once the operation base is established, then you can keep updating the information as being shared. Because in the cluster of the things, networking is a big factor. You know, it was never a big factor in Oracle. You know, if you're in Oracle DBA, you never used to worry about that. You know, now networking, you basically have, okay, even the 10G or 1G and one connection would make a difference. You know, and now you basically have Spark in memory things. You know, entirely different workloads are being done. So how do you do those? I think that's where we start with. And then we, these tools allow the vendors to see what's happening at the operational level. Because they're very good at technical problems solving. I mean, they can do, all the tasks can do that. But operation is as big a challenge on the day-to-day affairs in those enterprises. Abhimeda yesterday in the panel was, we were talking about ROI. Jeff put up some data that said, you know, people want to get a bigger ROI, but for every dollar they're spending, they're only getting 55 cents on their return. Because they don't have the skill sets and they don't have the knowledge. And then Abhimeda made the point, the ROI is not the way, what he's seeing is that the focus on ROI is not the traditional ROI, it's reduction of investment. Is really where they're getting the value today as opposed to sort of many organizations. What are you seeing there? No, you're exactly right. The, what we have been seeing in enterprises, the first thing that is happening is the cost reduction because if you're replacing, not even replacing, but you're not, no net new buying of data and the teaser of the mainframe maps. All that data. Baseline that. The baseline that. And that is moving all to Hadoop. And then you still, I think in the transition stage where you're performing the tarot datas of the world where you're doing the reporting and all the other things which have been doing. So the Hadoop hasn't been involved into that level. But I think with Impala and also the implementation which Hatoop has guys have been doing on the high side of the business and integrating the Tableau and the new visualization, I think you'll see the transition happening fast where they become sort of the EWs in the future. Well, so that, again we talked about this. There's a dollar that used to be spent on EDW and now there's 40 cents spent or maybe 30 cents. 30 cents, yeah. So been over here. So can they make it up in volume? It's like that story that I've been told yesterday. No, I think it's a challenge. I think tarot data has been smart. They basically are getting much deeper into Hadoop's face. And I think the markets are getting bigger. The provision of the data is so huge that I think they will sell licenses more cheaper but much more in numbers. Yeah, no, it's a serious question. I mean in concept they could make it up in volume. If you look at Hortonworks assumption that 50% of the world's data is going to be on Hadoop by whatever date they choose and I know that's sort of a marketing thing but okay, pick a date. It's going to happen. It doesn't happen, yeah. And with data growing like this and budgets not growing, the only way is to get on that curve and be a volume leader, right? No, no, you're right. I think the budgets will grow. Once the companies move beyond the cost savings, I think when the CMO starts seeing the value, I think the marketing budgets will grow. That'll basically affect the traditional, digital marketing and all those things. Those things are in the curve. But it's interesting your point about that and Abhi was very vocal about this yesterday. They're really starting with the dramatic cost savings. I guess it makes sense within the Fortune 1000. That's what they have to do. But if you look at these sort of born in the big data, born in the cloud startups, they're not worried about cost reduction. Everybody are creating massive value, Uber and the likes. And so, how do you discern in your experience, Manny, those Fortune 1000 companies that are going to be able to compete effectively? Because they are there. We know that some percentage of the Fortune 1000 maybe won't cross that chasm, but others will. What's the characteristic of that leader, that big data tech athlete that's going to actually be able to ride that digital fabric and compete with these new upstarts? Whether it's in retail or- I mean, so one example is that you basically saw as GE guys, how fast one can move as GE? I mean, they're offering a service now. So I think the Fortune 1000 enterprise have to do that, because it is an existential threat. We've seen HP dividing in two parts, semantic dividing in two parts. I mean, these things never used to happen. The consolidation used to be the thing. But they are getting nimble because they see the line over there. And I think enterprises basically look forward and kind of embrace this thing and take it on, I will survive. I think the enterprise, which won't, they won't. So appeal that a little bit. What do you see as the, maybe it's skill sets, characteristics. What is the company that can actually transform, look like? What should they be doing? What should they be focused on? What's the advice that you would give to them? I think the best thing what can you do is like, I remember when I came in my career, we used to be, Audically used to hire, IBM used to hire, and they used to give six months training to the support staff and everybody. And then they basically moved them to kind of move into the roles. Now that is not being done presently. I think the enterprises have to train the staff. You know, the staff has to take that to meet the challenge because net new numbers cannot come that high. You know, how many consultants they can hire? I mean, there are not many. So I think the existing staff has to be trained. You know, and I think the vendors had to basically do a much better job in providing online training and basically having training within the enterprises. So the existing staff can take on the challenge because until the existing staff doesn't take the challenge, the transformation doesn't happen, these things are not going to move fast. You know, from the bottoms up. And from the line of the business side is, I think they basically are making an effort into basically moving to those channels. And I think that will happen. So I think the will is there. I mean, that is the one different thing in this year is that there are serious business talks happening now. Which is earlier the hype, you know, for the last three, four years when you come to Strada. Now people are asking, I have this problem. How is it going to solve this problem? You know? As an SI, I mean, your whole life is solving complexity. Yeah. And it used to be in the old days of SI, you know, things like SAP implementations are so complex and it was a boon for the systems integrator business. But now what we've seen is things like, you know, the cloud and consolidation of infrastructure, conversion of infrastructure. It's sort of eaten away at that business and SIs really have to move fast. You're a specialist, you're born fast. So this question I'm asking is almost one of those, you know, it could, you know, put your existing business out of business, but what's next? But what are you as an integrator seeing in terms of the industry that the industry needs to do, gaps that they need to fill? Because who to do is complex? It's too complex. Complex. And then, and what does that mean for your business going forward? I think it means we have to be a continuous learning, evolving organization. Because what we saw in the market three years back at Astrolia Hadoop, I mean, till last year at Astrolia Hadoop, but we saw a lot of emergence happening in the new SQL space. So we moved fast and we basically started working in data stacks on Cassandra side. We won the big part, the first partner, one of them, you know, now we basically are the first SI to get certified on Spark. So what we are looking is that, okay, the change is going to happen. It's a constant change. We cannot rest. So what we had to do is basically we had to take all the knowledge which we learned from the industry, kind of like put together in the use case perspective and take these existing technologies that are coming at a much faster pace, have our staff kind of like have continuous learning. And I think the SI, that's a challenge. That you have to adapt. You have to continuously learn if you want to stay in the SI space. I think the SI's themselves are the first ones to go if they don't adapt. Other businesses come after us because we are there to solve the problem. If we cannot solve the problem, we are the first ones to go. What are your thoughts on the capital markets as an SI? We saw, I think, big analytics get acquired by TeraData. You're seeing hundreds of millions of dollars go into the big distro vendors. Service is not sexy, I always say. It's where the rubber meets the road in terms of business value. But from an investor standpoint, it's like there's other things I want to invest in. What's your take on that? I think it's changing. I mean, we talked about the 45% service share in the big data. It's a huge market. We are growing. We basically had 10 million this year. We expected to grow another 125% next year, $25 million. It's a serious business to us. We live through it and we are having fun of our life. So I think it's a serious business. I think it will drive the next generation of their big data technology services. And you see it from the fact that all the vendors are establishing a lot of service component to them. Yeah, I think it's critical for the technology providers to establish these relationships because as you said, if you're not able to move these things into production, that's where the revenue is gonna come for the technology providers. Yes, yes. So they need to make that happen and happen seamlessly as possible. I wonder if you could size up your competition a little bit on the big side. So how do you see the big guys of the world, the Accentures of the war, Deloitte's of the world? They're starting big data practices. You're seeing their marketing start to mention big data and analytics. Are they able to adapt? I mean, as quickly as somebody like Cloudwork who was really focused on this space? Well, I think they basically have a great, they have the domain knowledge and the relationship which exists in the market. But the challenge obviously for any big company is how to move fast into the space. And our focus has been on the not of the sexy part of the business and analytics part. We have always been focused on the data engineering part of the business, which we think is a very scalable model because everybody, before you do analytics, you had to do engineering. And I think in what we have finding in our space, they're not much competition. I mean, the field is so wide that you can keep, it's an execution challenge more than anything else. Well, what's strike mean? They may even be a potential partner for someone like you because they don't have that skill set. And they, but as you mentioned, they've got those relationships. And when you can tie your competency in the integration and really getting the system stood up and in production, and you tie that to the business use case and the business value and talking to the business side. And then I think there's potentially a win-win for both. I mean, I think that that's what we have been seeing. We don't go directly to the market a lot. We basically have partner aligned. We are really close to the light guys. We had an introduction to Cap Gemini guys. So I think things are happening in that area. They see the problem. And I think we have a solution for some part of the problem. I mean, they tend to compete in a couple of dimensions, which I presume you don't. I mean, they have like super deep vertical industry and micro vertical industry expertise around the globe. That's not your game. No, that's not our game. Your game is you're a dupe experts, right? I mean, you're that ecosystem, you know, players now, but you have to stay ahead of those guys, obviously. That is true. And so, so how do you do that? You got connections. You know, we were talking offline about the guys that went disco who were like super alpha geeks. You know, those kind of, you got those guys on staff? I mean, is that kind of? Yeah, we got these guys on stack who basically can go and evolve the technology, look at the technology. And we are coming a big part of the ecosystem now as it's grown because now it is not only, because the vendors are focusing on their technology. So if you look big data, it's getting complex. I mean, she talked about the Hadoop is complex. Eight different components. You put no sequel in it. Now you put spark with it. You know, it is becoming a horizontal stack in which there's not one player who saw the solutions. You know, so the stitching has to be done. You know, you have basically visualization aspect there. You have the log management aspect there. I mean, there's so many players in the industry right now which are providing a solution. So I think our value comes is that we can stitch all these things together on an engineering basis, you know, and to provide a complete solution where the vendors have maybe a large part of that but there is an associated ecosystem which has to come together too. Well, so in related to that, so we're seeing, I mean, just a flurry of, you know, partnership announcements between the technology providers. Now I'm certified on Clutter. I'm certified on Hortonworks to work together, whatever. How does that impact your business? Now, if the technology providers are integrating more closely together, what does that do to your value proposition? Or are you seeing, are we seeing a lot of quote-unquote Barney partnerships, you know, kind of just a logo to put up on a press release? I think it is to some extent. I mean, like that, but it is also real, you know. But you see a lot of these things happening. I mean, if you see, we were just talking to the elastic search guys. I mean, they have a big component of locks, locks-tash and locks-search, which you can do with the elastic search and come with a Splunk-like solution. You know, now they run basically on Hortonworks, they run on Clutter, they run on MapR. So, you know, those kind of things are valuable because they are part of the solution. I mean, the solar search is a big component, Lucidworks is doing something, but same Hadoop is doing, Clutter is doing on the search also. So those things are going to be there, you know, Strom, Kafka, SAMZA, now. I mean, you have all these names coming up, you know, and they are continuously evolving, you know. But the stitching has to be done. The underpinnings is the same from the engineering perspective, it's a distributed programming paradigm, you know. So once you get that, it becomes rather simple for you. Okay, Manny, we're going to have to leave it there. Thanks very much for coming to theCUBE. It was always a pleasure. It's great to see you again. Thanks a lot for having me, Jeff. It's amazing to watch the way in which this ecosystem is developing. You guys are right at the heart of it. So congratulations on all your success and what we're watching. All right, keep it right there, buddy. We'll be back right after this word. Thank you, guys. Great to see you.