 In fact, this is Dave Vellante, we're live with Silicon Angle and Wikibon's continuous coverage of Hadoop World 2011 and I'm here with my new co-host Jeff Kelly. Jeff, you've been checking out all the action today, so thanks for coming back and we're here, we got a great guest, Tasso Argeros, who's the VP of marketing and product management at Aster Data, welcome to theCUBE. Thank you Dave, first time on I believe. Yes. Yeah, so that's great, but it's been quite a year for Aster Data and the space in general, the enterprise data warehouse space, analytics and Hadoop and so you guys must be excited, congratulations on the recent acquisition. Thank you very much. And let's talk about Hadoop World, what's going on here for you guys, what do you see? And why don't you start by taking us through, sort of take us back to the acquisition, what's happened there and what's happened since and then we'll get into the Hadoop Angle. Yes, absolutely. So, the acquisition was a great thing for us, what happened was that we had, Aster was kind of the leader in this emerging technology space of big data for the enterprise, data that was obviously the leader in, data warehousing, enterprise data warehousing and so the merger of the two, I think created a leader in the big data space. So, the great thing about us being part of data is that we have a lot of, we have great amount of resources, we've been able to increase the amount of R&D investment that we're doing. We also have access to resources that we didn't have before. For instance, one of the things we did after the acquisition is that we came out with a hardware appliance that fully packages our software on a hardware packaging that's fully supported by data data and that's something that would be very hard for us to do frankly as an independent company. So, we're very actively exploiting synergies between data and Aster in terms of marketing, just technology and the amount of resources and support we have is great. Now, what's, did you plan on dropping Hadoop into that appliance? Is it already there or what's? No, so what is that? That's Aster software on an appliance, so it's our software on an appliance packaging and it also includes management software from data data and includes an Aster data connector. So, if you have a data warehouse, you can just drop this box, plug it in and very quickly get up and running and do ingest new types of data and move them over to your data warehouse after it's analyzed. Okay, so what about Hadoop World? Why are you here? What are you doing at the show? Well, I think Hadoop is obviously a great trend, it's a very big trend in the market. We do have, we're getting a lot of interest from people that are either using Hadoop or they're thinking about using Hadoop and most of the time the interest we're getting is that people love what Hadoop is promising, they love MapReduce, they want to be able to do deeper analytics. On the other hand, they're trying to figure out what is really a good use case for Hadoop, what is not a good use case and how can they reduce the overall total cost of ownership because the traditional way of deploying Hadoop which is you have a whole bunch of engineers and you deploy them to manage Hadoop to do the analytics, everything together, that's a model that works for a few companies like Facebook but has a very high overall cost. So what enterprise are thinking today is how can I deploy in MapReduce, how can I do this big data, new type of data processing but with a more controllable, more reasonable cost and that's where we can compliment Hadoop as a solution. So I talked to a number of Teradata customers, not just Teradata but other enterprise data warehousing customers and they describe what they go through as a sort of a snake swallowing a basketball and they're constantly ingesting more data and they can't keep up with it and that's one of the problems that Aster I guess set out to solve. Correct. And so I want you to talk about that a little bit and then talk about the unique requirements of Hadoop. So start there, tell people a little bit more about Aster data and how it's different as a platform and then we want to talk more about Hadoop. Yes, absolutely. So what Aster is doing is Aster offers implementation of MapReduce which is a language you also find in Hadoop and is the component Hadoop that allows you to do deeper analytics but we combine it with SQL. So we have a patented framework that's called SQL MapReduce that allow you to do MapReduce analytics but access those through a standard ANSI SQL interface. What this means is that business analyst and enterprises can do MapReduce analytics just like they could do SQL queries and SQL reports and you can also connect a lot of third-party ecosystem tools like BI tools, ETL tools to MapReduce through the Aster platform. So the way to think about Aster is that Hadoop is a great platform to ingest data, archive data and do batch transformations but Aster is a great platform to do a lot of iterative discovery analytics. If you have data scientists that need to quickly do a number of analytics, iterate through the data and expose those analytics to business analysts or BI tools, Aster is the only solution in the market that can do that. And in fact, we've gotten some tremendous traction, tremendous press around SQL MapReduce technology and even actually as I said this weekend there was an article in Forbes.com that was talking about technology and everything that you can check out. But really the combination of SQL and MapReduce is what it takes to bring the part of MapReduce to the enterprise. I wonder, are there any trade-offs you have to make when you wrap MapReduce in the SQL wrapper? I mean, do you lose any functionality, any capabilities? What's, are there trade-offs there? The way to think about it is that really what Hadoop has optimized for is scalability. Right, so Hadoop is super scalable, you can have a Hadoop class of like 5,000 nodes. However, for the people that do not need a 5,000 node system, we have traded some scalability for a lot of performance, ease of use, enterprise readiness. And what we find that for 99% of the customers out there that is a fair trade-off. Also, what SQL really gives you, the ability to iterate very, very quickly, it's very friendly to analysts, very friendly to data scientists. If what you have is engineers that all you want to do is just write code, code, code, Hadoop is obviously a great platform for that because it is a development platform at the end of the day. But most enterprises though, they're not willing to make the investment and hire an army of developers just to write code. So a year ago we were here and there was really one player, Rose Cloud there, and now you've got others entering the market. They were there, but a light has been shined on this. What's your take on all that? Is it a great thing because everybody just gets more business? Well, competition is always a great thing, at least for the customers. I think what's happening is that it's very simple, the Hadoop market is growing and there's more and more people that are, you know, willing to go out there and offer services and offer support and even offer products around Hadoop. From our perspective, we believe very fundamentally that our value proposition is very unique. So the more people can offer Hadoop services, the faster the Hadoop market will grow and the better it's going to be for us because we have a great story about how we complement Hadoop to enable more analytics for enterprise users. What's the big gate in your view? I mean, it's obviously growing. We saw this show was probably 800, 900 last year. It's, you know, I think Mike said 1400 this year and probably, you know, growing at the pace of the community. But what do you see is the gate? Is it skill sets around data scientists? Is it the maturity of the technology? Is it the ecosystem, you know, all of the above? Can you talk a little bit about that? Yes, I think it's definitely a combination of things. I think there are surveys out there, excuse me, that you can take a look and what they say is that a lot of enterprises getting stuck either on the skill set that they need to have to, you know, roll out these Hadoop deployments or on the connectivity with their existing investments and ecosystem tools. And this is something that Astro can really help because we can mitigate some of the skills that are needed. We can move some of the skills that need more to analysts than developers and we can also help connect MapReduce to ecosystem tools. But it's definitely a combination of ecosystem connectivity and skill set. I mean, if you think about it, you know, how did, you know, place like, you know, Facebook and Yahoo and place like that, how did they become successful with Hadoop? And, you know, Google, which is really where MapReduce started, right? How is Google successful with the MapReduce technology? They have very smart engineers that can be an analyst, a developer, and a system administrator all at once, right? And that's the traditional model of how you make a Hadoop deployment successful. You hire people that can do all three at once. They can figure out what's the business problem, what it means in terms of analytics, they can code up the analytics in code, and they can deploy the code and they can run it. They can do what, you know, what is called end jobs. So this model is clearly unsustainable for enterprises. You know, if you Google and you write next to Stanford and you hire all the species, that's great. If not, you know, tough luck. So I think you will see the market evolving towards a place where Hadoop will have some specific use cases that it's going to be extremely successful on. But then you have other products like Aster and others that can serve other use cases that Hadoop cannot solve today. And I think we're looking at the market moving towards the direction very quickly. Do you, let's just listen to you speak to us so when I see a services opportunity there, are you guys going after that opportunity or leaving that to others or partners? So right now, the way Aster was developed is that we do have services. We're not, services is not our focus, so we partner with a lot of SIS in the market to help deliver those services, but I completely agree that services is a big part of the story. One of the great things about becoming part of TheraData is that TheraData does have a very substantial presence of services. So we now are able to offer a more full solution to our customers than what we could offer as a small startup with very limited services resources. Excellent. Well listen, thanks very much for coming on theCUBE today and sharing with us some of the perspectives, the acquisition, congratulations again. Sounds like we're in the early days of Hadoop, but you guys want to be there, you want to play and you got a lot to add with a big new parent and so well done and we'll be watching. Thank you very much. Glad to be here. Thank you.