 at Big Data SV 2014 is brought to you by headline sponsors WAN Disco. We make Hadoop Invincible and Actian Accelerating Big Data 2.0. Okay welcome back everyone we're live here. We're here for Big Data SV, we're covering all the action here live. All the thought leaders, entrepreneurs, CEOs, venture capitalists, also covering the Stratoconfer's going on right behind us at the Santa Clara Convention Center. I'm John Furrier, the founder of Silicon Angle. This is the Cube. I'm joined by my co-host Dave Vellante, co-founder of wikibon.org. Our next guest is Michael Sandoval, the CEO of a TGO. Welcome to the Cube. Thank you. So welcome on. Tell us first about the company and share the folks. You're from Washington, Seattle area. Talk about the company and what you guys do. Sure. So first of all we're Seahawk fans of course. Seattle you know is clearly the second Silicon Valley if you will. So a lot of our former Microsoft folks we founded a teaser back in 2005 and the idea was how do you take and tap into the exponential growth on data given all the heterogeneous environments and so we decided to kick a company off that really attacked that from a platform perspective. Talk about how old you guys are when you started. What's the backstory on the company? Yeah so 2005 is when we kicked the company off. So we've been around a while and it was really the long-term approach to how you change the 21st century big data. You know I think the challenge is a lot of companies are coming out and you know they're new to the big data you know arena. So the thinking about things in a perspective of applications on Hadoop or attack the specific problems. We really took the long road five to ten year plan on building a platform that could go into heterogeneous environments and extract out the data leaving everything where it is and then actually realize the patterns out of that data rather than trying to look for specific patterns. So it's a different approach. You know Dave and I were talking on the intro about the comparisons of web 2.0 to big data the market I'll see there's some trigger points in between one early on away great recession but it's really a bigger than that one there's some business model opportunities one with the growth and two it's hard right so you can't just whip up a start up and jump into the in these platform games because it's pretty complicated. Some cases maybe if you're doing you know maybe the point solution but to build the platform in that space is pretty difficult. Could you just share your perspective of some of the things of why it's hard. What are some of the hard problems you guys are solving and what's the benefits to those. Sure so when we look at the you know the data across the world clearly your data sits in its native states and everyone you know in the 20th century was using specific tools and I knew when I was a young kid my father said you know use the right tool for the right job and the problem is a lot of the problem is today a lot of these companies are you know taking specific data mining approaches or machine learning approaches or all this typical 20th century approaches with statistical you know methodologies basing in statistics or whatever I can go through all the technology but the reality is all of our customers think of data as very specific and valuable to their business so how do I take and approach that for each and every customer that doesn't take a bunch of professional services so you got to find a way to extract the the meaning out of the data for the specific application for the specific customer if you start with a problem first you have a small set of customers you can go attach to so we have to take a approach that you get in there quickly and realizes the value out of the customer's data and then help them actually discover the value in their data so it's basically what you're saying is it's a hard problem is you can't just go in and propose something new that's going to be foreign to these guys you have to go into their legacy environments and adapt absolutely you can't go with a hundred million dollar proposal to spend you know 50 guys in 18 months to do professional services you know activity to aggregate their data some of the clients in the market all these other scenarios that you typically see some people are still hanging out of those days so now you guys are a platform play right you partner with acting can you describe sort of what you actually what the solution actually looks like sure so when we think about the platform there's lots of layers to how you take in and approach the marketplace so when people talk platforms often they think in terms of do I have restful API's can I expose it for people to build applications off of that you have to go even deeper when you're talking about the data residing in native you know states and data silos you've got to attach to that in the infrastructure layer so when you have great in infrastructure analytic companies like Acton you know they're going out and they're finding new ways to connect to that data to use that data and we're you know doing some of that and adding lots of connections in that as well but we really focus on the analytics layer so what can we do in a hybrid system across all types of analytics whether it be you know traditional machine learning natural language processing inferencing so deductive and inductive processes you go on and on about the diverse approaches but that's what we do and then we together build this vertical platform if you want to call it that that exposes all the application availability for all of our different customers in different industries so we've got people in government we've got people in health care you've got people in retail consumer marketplaces we have to be able to attach to all those forms of data and in those heterogeneous environments so do you go after specific use cases in specific industries or as a platform player you more opportunistic yeah we we have focused to quite a bit on the health care space and in last 12 to 18 months so in that space there's significant problems whether it be the transition from ICD-9 to ICD-10 which a lot of the payers and providers are mandated to do this calendar year October 2014 and that's shifted twice because a lot of companies were out there helping them to to realize this 20,000 codes to 155,000 codes they haven't had the appropriate technologies to actually migrate these you know health care providers and payers so we are sorry to interrupt that's a that's a is that a compliance use case or is it more of a sort of execution and be able to right it's a little of all all of the above you have clearly you know hip a compliant data that you have to attach to you have standards and and and law that's requiring you to code the encounters with doctors in this new in this new ICD-10 international coding standards so when you look at the compliance meets all the regulatory you know rule set and you look at all the various data from pharmacies to payers to providers to dentists all of these people are trying to trans you know transition their data and it's not as simple as let's say moving my paper files my and using OCR technologies to move to a EMR you've got to use something that can actually take action on that data so when you look at unstructured data how do you do that when you have different disparate data systems and all the healthcare you know providers you know hospitals and clinics and small dock offices how am I going to access that when they don't even know and their IT people don't know where the data sits and so you've got to find a quick way to go across to all that and expose it so there's lots of healthcare examples there's lots of examples and financial services you know Dodd-Frank and and compliance scenarios like that that we're going after as well so that's as well and I would imagine back to healthcare for a minute that the EMR piece there's probably a meaningful use component of the ROI that if you could show meaningful use you're actually going to get paid right absolutely so when you think think about you know population health you think about your things hopefully transitioning 2020 to individualized healthcare and you look at ACO measures to you know scenarios around you know what's really going on in the administration and the readmittance scenarios that are going on right now so it's not as simple as checking the box and saying did a patient have congested heart failure and are they over this this age there's all these symptoms that may be in the infrastructure data and how do you realize those actually take and use as a signal and a predictive measure in order to in real time you know realize when a patient is checking in that they may have a problem and may may readmit if I don't walk walk through to a different workflow than the general populace Michael talk about you mentioned consulting before that people don't want a hundred million dollar proposal you've been in that industry you've had a lot of space of Microsoft say large large experience in the enterprise and deal with customers so two questions one is David I always love the services angle and how disruptive the new technologies can either shorten the cycles get the time to value faster so the question is what things you see on the time to value in this new market with big data and on the deployment side in terms of acceleration the cycle of concept to deployment and two is it truly vertical where everyone talks about big data being a very vertical approach where in the old days you have to carve out a vertical and go deep and be you know kind of focused on one vertical where you start to see vertical solutions with kind of a horizontal platform it seems to be what you guys do so comment on the time to value cycles of you know services delivery implementation and then talk about the vertical impact is that still a mega trend with the vertical focus those domain experts those are really two big questions yeah I will say unfortunate unfortunately it is a trend to go deep in vertical and I think the problem then is and I've had one of our government clients describe it as you know a bodybuilder scenario you know big body application on little legs to do and so a lot of scenarios like that with lots of professional services using using those old scenarios of I'm gonna throw you lots of people will figure out how to customize this and and be in your business all day long for the next few years I think now when you look at the cost effectiveness of you know cloud-based technologies and platforms to be able to go across the data and realize the patterns out of that data you know more more automatically we are finding scenarios in one of our clients MD on quoted publicly last year they process two-thirds of the healthcare claims in the private space in the US and when you look at at their scenario a lot of the big boys that you mentioned earlier that typically go in professional services in those deep you know analytic solution you know verticals they couldn't get there in a long period of time we got in there in a matter of a week and realize the data for what they were trying to accomplish and their quote was we did in 130th the time and 130th the cost and I think the other quote was before others could even come up with a plan and so I think it's a brave new world to the 21st century is about using computers and software in in any type of environment to realize the getting to the value you guys got to the value while everyone's kind of eventually rearranging the deck chairs on the proposal you guys essentially kind of chop it down into little chunks is that kind of you see it yeah I don't know about shopping around down a little chunks you're trying to connect all the data and depending what those chunks may be you don't want to have to use a bunch of professional services people in order to in order to you know figure out which hypotheses and what questions you're going to step in first then ETL the data and then go oh gee now we got this new data can you add that to that yeah a month later we'll go through this whole process yet again you want to do that in real time you want to be able to when you look at predictive measures you want to build this in forecasting but you also want to realize what's going on at the moment in time in real time so some will argue that's a great business model professional services create a bigger problem that you come back in and and solve which is kind of an old model and that's why we you know we talked about this notion of disrupting the old model which is you know getting to the value faster versus being dependent upon you know the the professional services drug and so that's interesting when you look at the research that we've done a wikibon at the still the bulk of spending in big data goes to services so Michael I want to ask you you hear a lot about you know these conferences like Strata about being a data-driven organization and you know data is the new oil these bromide what are you seeing in the customer base our customers truly transforming to be data-driven what does that mean to organizations and and how specifically are you helping them get there yeah a lot of the customers say that you know they have so much value in their data and I ask what are you doing with your data well we can't really access it right now so they become data hoarders so so they are data-driven and when we look at our mission which is compassion technology for a wiser planet and tying back to John's earlier question is you want people to make the wise decisions on the data you want the data speak for itself to expose itself and use it to actually you do more with less with the people if you will so how could I take the same professional services where I put 50 people into an account and put two people in that account and take the 48 and put them across a number of other accounts I think that's where we need to go so the companies are data-driven today they realize that they have to change they fundamentally have to shift to using their data but also what data might give them lift or optimize their data so if I can realize those patterns out of that not just my hypothesis or your hypothesis in being in the business but what does the data speak and what other information that might be in the marketplace that actually can edge on to that they give us more power in the marketplace and that is really becoming a data driven well I remember you know the Harvard Business Review articles in the sort of early 2000s about you know gut feel oftentimes you know beats you know Trump's data and that that is flipped I mean I think in general but at the same time for an organization to become data-driven they've got to do some things at the back end it's not just technology right it's other things what are you seeing in advising your clients in terms of organizational changes that they have to make are they putting a data czar in charge like a chief data officer I wonder if you could talk about that a little bit well I think the the the owners of the data in the past have been the IT staff and so how do we take and and shift that to the line of business people who are actually trying to take action on the business and so is it really in your best interest to try to you build a large IT staff and and use them to you know govern and help you through the hypothesis process to discover what's in your data and what how your business operates so I'd like to see those resources shift to the business side of the equation and maybe shift all that IT out to the cloud and something that can actually be you know actionable here and now so when you talk about big data being you know you know fundamental problems in volume and variety and velocity the cloud if I can spin up a thousand servers to go take action on this new analytical exercise I can actually do something that can affect my business today rather than taking a bunch of people IT order a bunch of servers and go through the traditional processes so that that is shifting John I talk a lot about the this role of the chief data officer we have a partnership with the folks at the MIT information quality group and they have a conference every year on the on the CDO the chief data officer and they've put forth the premise that the CDO should not report into the CIO it's really a parallel type of path independent type of path in particular health care government financial services some of the areas that you're in a seem to be getting momentum around that that thinking I wonder if you're seeing something similar or is it sort of early days are you seeing real examples of data czar as a chief data officers emerging running parallel to to we're not even parallel but independent of IT we are we're seeing a lot of transition I want to say a lot a new transition in let's say 10% of our clients that are going to those type of scenarios there are more they're empowering the the you know the former business leads the VPs of you know operations and sales and what not to become those data czar's rather than leaving all that power in the CIO's hands unfortunately I think there are a lot of customers still of of the large professional services organizations that still believe in the old philosophies and I think that will shift over the next 10 years now that brings up a security question how does the security model change when you start thinking about data being the main driver when you think about platform to help me address some of my data challenges what are your thoughts on security would you guys you know doing there what's your angle and approach to security clearly fundamentally security and privacy have to be built into your platform from the get-go it can't be an afterthought so it's not an add-on function or feature that you turn on and off it's something has to be fundamental to your to your data process so so we are including that within the platform as is acting and together we're able to deliver that from infrastructure all the way through the applications so in the last couple of months we did 10 billion HIPAA compliant transactions in the cloud and a lot of the CIOs go well you know we can't do our compliance we can't do our you know our HIPAA transactions in the cloud we can't move to the cloud for this reason or that reason then you could fundamentally work with them for maybe a couple of months to get them over those you know 20th century prejudices and realize that actually you can do this and maybe even bet better than the traditional security measures they have within their own infrastructure so how do you guys don't if you can talk about the way you compete we've talked about a platform we talked about the partnership with acting how do you compete in the marketplace what's your unique advantage is it a case where you're sort of coming in with a platform and others are coming in with point products and you're differentiating that way I'm sure that's part of it but I wonder if you could talk about that a little bit specifically in terms of you know when you win why do you win right well I would say that as I talked a little bit earlier a lot of the big data companies that are out there are solving specific you know one-off problems and they're and they're really good at those problems they're they're going out and solving something specific for financial services something specific for government something specific for health care but when you have a platform you have to be did agnostic you have to go into any hitter genes environment and be able to expose those patterns so our unique advantage is to help those companies actually get a foundation under them that it deals with privacy deals with security reduces their their you know need to use professional services when they go into sort of organizations and conversely when you look at acting on ourselves where they go in across the infrastructure and attaching this data and doing the you know various analytics and we go on top of that to do you know deep analytics and expose this for application vendors and big data that kind of relationship is is really how we're piggybacking it so we're building solutions with a lot of partners even even some of the the likely suspects and big data that are mainly professional services now are looking to us to use the platform to get more out of the professional services people and win these deals that have repeatable technology so sort of new new new alliances and partners like a final question for you as we wrap up here given all your experience you've had in the industry we kind of pinch ourselves we've been around so long like we've seen the cycles before you've been at Microsoft the growth has been well documented companies story history and now as new leadership what is your vision for the future why are we such an interesting time what is the inflection point why is it so publicious why is it so hot there's a lot of growth that we were coming in the early segment give your personal perspective on why this time of these trends can be why is it so important why is there so much disruption why is there so much activity right so you know we came out of the 20th century with a big industry around computers and software and that was fabulous and relational databases took us so far we all know the exponential growth of data and you could talk about that and how it's impacting things but the reality is and we've all heard the quotes you know in the between the beginning of time and the end 2000 in the next you know a couple days will generate more data than you know all those years combined right so what is the real challenge the real challenge is that we have a growth of people we have a growth of industry around the world and lots of new ideas that leverage you know information and the challenge with generating data data is not necessarily information if you can't make use of it it's not information it's not knowledge it's not wisdom so all these you know talks about big data and the challenges with that is how do you go into any environment no matter what language no matter what data type no matter what protocol what no no matter what the infrastructure may be and expose and give an advantage to optimize that specific business and that's the challenge and that's where we're going with it with the 2020 plan if you will for all the companies out there in the exhibit hall strata you know there's always you know the startups are there you know and some will win some will lose it's kind of like the you know that expression when you go to college look to your left look to your right those people might not be there for the for the competitive nature of this market what is what is what does it mean to be successful what do these guys have to do to stay around some will tap out and and drop out of the race if you will in this market valuations are high the stakes are high great opportunity in the growth what do they need to do to be successful what does the company to do to really go the long haul well much like acting and let's use you coming together you need a team this is a big world a big data space and it takes more than one company to go solve these problems so that's my first piece of advice second is when you look at all the various analysts reports that are out there most of the the companies are tackles from professional services which is not necessarily repeatable so you need to take a lot of that knowledge and put it into your technology and your software and differentiate yourselves and so that's I think a key advantage and those who will succeed you know become the oracle if you will of the 21st century Michael thanks for joining us inside the queue we love the big idea is this what the queue is all about and we'd like to stream those ideas and knowledge if it's not being used it's not a knowledge the data is not being used as critical critical opportunity the growth is there great valuations great market opportunity this is the cube right back with our next guest after the short break