 Okay, we're back with SiliconANGLE.com's theCUBE, our flagship program. We've got the event to extract the signal from the noise. I'm John Furrier, the founder of SiliconANGLE. We're here at Strata Conference in Silicon Valley, and I'm here with Bruno Aziza. That's correct. From C-Sense. C-Sense. C-Sense. C-Sense. C-Sense. The starters, so you guys have news today. You're on Venture Beats front page. Congratulations. Thank you. Welcome to theCUBE again last year at Strata. Only a few months ago you had the tats on and doing the awareness. So let's give the update. What's going on with the company? Tell us what's happening real quick and then we'll jump in to quick commentary. That sounds good. So there's no tattoos this time, at least the non-places that you can see. So since then, at Strata, we talked about one terabyte on the $1,000 machine. And we're a new startup. We've been in the market about 18 months now. We've got about 400 plus customers. And so really the news has been going out and it's been growing like crazy. And of course you've read a lot of the coverage, but we've also acquired lots of new customers and lots of people that are interested in finding out why is this technology different? What's different about it and what can I do that I couldn't do with earlier big data analytics solutions? Great, so let me ask you a question first before we get into it. What do you think about Strata this year? What's your take on so far? Obviously it's day one, it's the quiet day. Tomorrow's going to be a bigger day. But obviously there's some big moves by the big money players, EMC. So what's your take on what EMC's doing? Well, I don't know that I should comment on EMC in particular but what I do see, you know, because I'm not informed enough about what their strategy is and I'd rather talk about what I'm working on. But what is interesting about the movement in general is it's growing like crazy. I mean, I've always been a believer in Strata, the movement, I think it's a different conference. You see lots of people that are really deep in the technology and we need that. Last night I was at the end conference so I did a little lightning talk and my CTO had a 45 minute session that we, and you could see how much people really get the technology and they're eager to understand what we can do now with hardware evolution and software. It seems like in the past we've talked a lot about software evolution and now people are getting, you know, starting to understand that hardware evolution particularly on CPU and memory hierarchies, you can crunch a lot more data that you couldn't have passed so that's encouraging. In terms of the major trends that you're seeing out there that affect your business, what are they? So there's a few trends. The first one is what I call the commoditization of big data and by that I mean that any person is now is going to be running with terabytes of data on their machine, on their local server and so we have lots and lots of data that the average person has access to and so that's number one and today there are a lot of user friendly tools that allow people to work with that. The second one is what I call consumerization of big, so one of them is commoditization and the other one is consumerization which is customers are asking that they get access to software, enterprise software that is more like consumer grade and so what that means is in our case, for instance, they want software that does everything from A to Z. You know, today when you work with big data you have to buy a database, you have to buy an ETL platform, you have to buy visualization tools, that's expensive, it's complicated, it takes time. Customers don't have the tolerance for that and that's why, you know, I think we're getting a lot of success is we're optimized for their hardware, it's one solution for the database ETL and visualization and so it makes it super easy for them to get started and that's really what we need to do next is it's a big data thing, we've been talking about it but we've got to be careful that it's not just for the experts and the few, it has to be for the 600 million users out there that are dealing with data on the databases. So obviously on Twitter we're watching all the traction you have with your demo. Yeah, it's a record, yep. C-Sense, 10 by 10 by 10 challenge, Shatters records at strata, big data conference, analyzes 10 terabytes of data in 10 seconds on a $10,000 machine. Okay, so just take us, that's impressive, congratulations. What does that mean? What does that, but what kind of benchmark are you running? Because remember, Green Plums making a lot of claims and 100 times performance over Impala, but Impala is 100 times more performance over Hadoop. So I'm just, people want to know the math. So the benchmark that we're measuring ourselves against is, okay, so now you're in a situation as a customer and you have terabytes of data. And we're not talking pitabytes here, we're just talking about the most common situation that customers have, which is terabytes of data. And today the solution is you need probably 20, 40 Hadoop nodes to equate this performance. And you might not actually be able to produce the- What's the benchmark on the data? What kind of data is it? So it's data coming from multiple data sources, you know that- Structured data? No, it's structured and unstructured data. It's a little old. So it's a little bit of both, you know, if you look at our technology, we connect to about 80% of the world's data. So it could be Hive, it could be as file. Yeah, because you guys are pulling in multiple sources. You're doing like the data mashup. That's the number one- So you're 10 terabytes, just before we get into the weeds and the math. The 10 terabytes is mix of different types of data. Correct. Okay, not one- No, no, one simple file- No transformation, SQL boosted up. No, no, no, no, no. This is real workload type data. That's workload that's a real life scenario that a customer would run into. So, ring 10 terabytes. Of course, we have our technology offers compression and we do optimization of the data. So by the time it gets to the analysis phase, it's compressed, right? So it's a lot smaller by the time we crunch it. That's why we can crunch much faster. But what we're trying to replicate is a typical scenario that a customer would experience. So the first piece of the technology is this high performance analytical database we call the ElastiCube, that's part of the solution. The second piece this solution does is it brings in the data and does automatic ETL. So we detect the relationship between tables, we do that automatically. So the end user doesn't even have to understand what a join is. That's the idea. And then the additional value on top of that is that we have visualizations. They're web-based, HTML5, JavaScript. You can also extend that with D3 libraries. And once you build these dashboards on top of this data, you can now query this at amazing pace. You know, one of the things that people don't think about when they design dashboards on large datasets like that is that the typical dashboard has five, six different queries running at the same time. So when you look at other vendors and you say, yeah, I can run this in three seconds. Yeah, you can do that on one query. But what happens if you have 500 users running five queries at the same time? Most tools choke. And what we're trying to show is that on one Dell server you can do this and never choke. Got it. So okay, so now let's take that to one step further. How do you translate that out to scale? So in your conversations with customers, so it's a great demo, shows the speeds and the speeds like doing a lap around the track. One lap really fast. What, how does that scale into, what are you guys seeing in terms of some of the concepts with the customers? So most of the customers we're talking with, it's actually interesting. They don't think about the, they don't measure their data by saying, well it's five billion rows and 10 terabytes only. The problem that they have is they actually don't know how quickly the data is going to grow. And so what we tell them is like, think about it differently. We're going to take a look at your entire data set. We're not going to sample your data. We're not going to try and limit your analysis. What we want to do is we want you to not even have to think about how quickly your data is going to grow and have a solution that's going to scale with you. So this is one node. You could have multiple nodes just like this one and scale to unlimited environments if you wanted to. But that's the mega point here is it's simple to deploy. It's one solution that does everything from the database to the visualization. And it's something where you don't have to worry about data growth, which in traditional deployments, that's the first thing people ask you. And if you went to most of the vendors out there and you said, look, I've got this data. What's the first thing they'll tell you? Give me a sample. We don't tell our customers, give me a sample. We tell them, give me your entire data. You know, why would you have a bottleneck at the analysis level? And that's our approach. So let's talk about you and the company right now. Because obviously when we first met and you moved out from Microsoft, you're very dynamic. You're very great. Thank you. I'm passionate. I'm passionate. It's the passion's key and you're doing a lot of stuff at scale, which is kind of my test for someone doing some cool stuff. So tell us about some of the things that have happened for you guys in the past. We saw you at Strata last year, you get a lot of awareness. You've got the big project here. What big things have you guys done and what have you seen out in the marketplace that has got you fired up? So what you're doing and then something not you're doing that you're excited about in the market. Yeah, so there's been a few things. First of all, I think when we saw each other in New York, we were just barely talking about what's happening. I mean, we had done zero marketing. And so since then, you know, I mean, I can't give you numbers on traffic and things like that, but it's been, we've had amazing demand in terms of customers, interest people coming to us. You know, I think when you start a company like this, most people are trying to understand, well, first, what do you do and what do I come to you for? And so now we have a much larger level of awareness from customers coming in and we have even people from the industry coming to ask us what our opinion is on the industry. So I think it's a great sign that we're now doing a good job at explaining what we do and the types of situations we help customers with. We've seen lots of very cool stuff in multiple areas. I think last time we met, I was telling you about how large companies are deploying the software. And we actually now start seeing people all over the, the, the gap. I mean, you've got people like Target that are using it to do theft detection. So they're mashing up data and they're trying to figure out, you know, theft issues and so forth. And you've got smaller companies like Wix or Plastic Jungle Wix is a website platform. They have 29 million users and they don't have a 100 people data team. They have like three guys and they're able to use their software to crunch through this massive amount of data where they do behavior analytics to improve their, their software. I think that's really cool. That's what I see the most is that people now starting to understand that there's two things going on. One is data is their products. So when you're Wix, it's the data that you have about your users that are going to help you build a better product and you have to really be able to crunch that in real time because the slower you are, the more the competition is going to catch up on you and they're going to know better how users are using this data. And then secondly, we have people like Wi-Fi as a company that does mobile hotspots. They have millions of users out there and they can tell you the status of a phone in a particular area and so forth. And then user software to crunch through large amounts of data to go back to the telco providers and tell them this is the quality in this particular area and so forth. So real time scenarios, large scale and small teams which is, I think is great. I think it's a great sign that we're evolving to making this thing really complicated to something that anybody can do. That's really the opportunity. And I think the interesting thing is the small teams. I mean, you're seeing kind of agile kind of mindset come into a lot of the big data space where we were at the Green Plum event and Harper from the Obama Industry Academy. His big thing was team data software. And I like that concept because it kind of works and that's the most effective. When you start to get the bigger teams, bureaucracy, agendas start kicking in. So I like that. I think I would agree with you but I want to ask you also about your view of the cloud because I think what you guys are doing that's interesting that I like is I've always been impressed with the idea that you're bringing in multiple data sets which gives you flexibility. You're not locked into say a data warehousing model where hey, I have certain queries. I'm going to run these sequels. I want to make them run fast. That's an old paradigm. The new paradigm is spreading resources around the network, having a data platform but dealing with multiple data sources and Intel was talking about that today. So when you talk to customers, is it a foreign concept? When you talk about multiple data sets, are they like, hey, welcome to thank you for coming? We need you. No, there has been a disconnect in the past where the older generations assume that, well, you've got multiple data sources. By the way, the average company has six to 10 different data sources. So every company you talk to, you have a Google spreadsheet and some other data source you're just dying to be able to match up. You can't do it. And so I think what you're describing is mainstream. The issue has been that most vendors have approached customers and have told them, no problem, the way you solve that is you call IT, build the data warehouse and we'll talk to you in 18 months so we can analyze this data. That is not the reality. Customers do not tolerate this at all. What they want is I just want to download something and magically my data works for me. And so that's the scenario we walk into. I think the other evolution that's happening is I agree with your comment on hardware, I'm sorry, on software. I think it's also the evolution of hardware and how it plays with software. So obviously in memory, in 64 bit, has been in great evolution. But now what's going on is that people are actually realizing, you know what, RAM is actually slow. Why? Because CPU and L1 cache, I don't want to get too little and will provide much bigger performance improvements. And that's all in your machine. I mean, the environment, the reason why we showed this is because most people running software on environments today, they are under utilizing the power that they have in their own hands. And we need to stop that because otherwise you're going to be spilling hundreds of servers and you're going to be using 10% of capacity. I equate this to the brain. We have this gigantic brain, 400 billion data sets that we can, but in the end, our brain looks like this. We're using 10% of it because we don't have the analytics for it. And that will change that. Well, I mean, we're big fans of startup, especially ones that are kind of breaking the barrier in terms of value proposition, bringing value to customers. Cloud-based products and services are obviously emerging. I didn't tell you my opinion on cloud. I want to get that as a final question to you because obviously the future is going to be and the puck is going to where we're skating, so to speak, to cloud-based and services that create business value. So not just make the horse and buggy go faster, that's the data warehousing model. Put some cheap dupe on it and you have cheap data warehousing. Everyone's going towards a platform for data. And so business value is the conversation. Yeah, they want the speeds now, but they don't want to foreclose that business value. So final comment on cloud and services. So there are two things that we see on cloud. The first one is data in the cloud. And I think you talked about that. Primarily, I think it's going to be driven by people that have applications in the cloud already. So if you have Salesforce, you have Google, that's easy. Porting your on-prem data into the cloud, that is a nightmare. It costs you more money and it's difficult to convince customers because typically the data that they have on-prem is business critical and the reason why they don't want to move it to the cloud is because it is business critical. So you end up moving data in the cloud that is not business critical, which I don't think you want to be in that situation. That's comment number one on the data in the cloud. Second comment of data infrastructure in the cloud, I think that's a big play because again, it simplifies how people get started with this field. We just announced so our software can be deployed on-prem, on Amazon cloud and we just added Windows Azure cloud. And the reason for that is because, you know, our mission is how do we make this super easy? How do you provision a machine out there, install the software and just get started? And so I think infrastructure in the cloud has a much bigger potential than actually data in the cloud by itself. Unless you're talking about open data or data markets. Well, no, moving data around is expensive, right? So and transforming of data is expensive. So I think you got the right approach. Bruno, we're getting the time hook here. Thanks so much for coming on theCUBE. Is it C-sense? C-sense? C-sense. I always get that wrong. It's an East Coast thing. Our product is Prism. Maybe that's easier to pronounce. I love the elastic cube because we are the cube. This is still going to angle to the cube. We are elastic. This is Silicon Angles coverage of Strata conference. We'll be right back with our next guest after this short break. Stay tuned. Thank you.