 from San Jose in the heart of Silicon Valley. It's theCUBE covering Big Data SV 2016. Now your host, John Furrier and Peter Burris. Okay, welcome back everyone. We are here live in Silicon Valley in San Jose for Big Data SV, Big Data Week, Strata Hadoop, all the actions happening here. This is theCUBE's Silicon Angles flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier with my co-host Peter Burris. Our next guest is Jerry Held, board member of so many companies I can't list, but the notables are Informatica, MemSQL, Tamer, NetApp, Informatica, a few more too as well. Welcome to theCUBE. Great to be here. So exciting to have you on because we were talking before we went on is just a history of what you've been involved in. I'll see with Mike Stonebreaker as his first PhD student, Ingress database all the way through a story career. And it's been fantastic and it's great to have you here. What's your take right now? Looking down on the balcony from the balcony to the stage of Big Data. Are we seeing the same movie again? Replayed over as a Groundhog's Day? Is it the same movie over and over again? What's new and what's your take on what's happening right now? What is the big aha moment that you see? Well, over the four or four and a half decades that I've been involved with this data stuff, there's these transitions. Every once in a while, you get a big transition. I started Key in an O29 key punches with paper cards and then we moved to online systems. Tandem, we did the fault tolerant systems and we want to put the whole industry online. PCs came along, client server. We've gone through all these transitions to the web, web-based applications, the smartphones. Each transition that comes along is both an opportunity and a threat both for users and for vendors. And if you look at each of those transitions, successful companies have died, new companies have formed, and some of those have come on to be big winners. Did they make the next transition, the next transition? Typically not, some do, most don't. Right now, we're in the midst of really significant transitions on multiple fronts and we can talk about some of them, but where data is stored? We're moving from on-prem to the cloud. We're moving data in from all different sources. We're doing big transitions on how data is processed and we can talk about the new forms of processing of data. And I think most importantly, it's a whole new level on how we manage data and how we organize it and how we take responsibility for the data in a corporation which is becoming the most important asset. So all of those are big transitions that we're going through now. Similar to the past, create great opportunity but great opportunity also to lose money as well. So where the data is, how it's processed in management, okay, sounds pretty straightforward, but I wanna get your take. I mean, you mentioned some things that are going on and each one of those theaters of transition is massive. Compare that to other transitions in the cycles of the innovation of the technology industry and how does it rank? I mean, it seems to be bigger than ever, it seems to be combination of some of the biggest inflection points we've lived through. How would you put a color around this transition to scope and size and impact? Each transition is critical. Each transition is critical in terms of customers adopting the technology and making their business more efficient. Each transition is critical for the vendors involved. Is it bigger or smaller? The whole industry obviously has grown. But take one of the very first transitions when the relational database industry was born. I was there, I was actually in the delivery room. And if you can remember back then, there was one company that dominated this industry completely on every aspect. IBM owned every aspect of computing, including IMS, which was owned the data, all of it. And they still exist, they're still a very successful company, but their 100% market share went down dramatically as companies like Oracle and Sybase and others came along and we moved to the new relational model, which has survived a long time. But those transition points have had percentage-wise big impacts. Right now we're going through huge transition. And the thing that's different now, I think, than then is in those days, companies automated only a small part of their business. In the 80s and 90s, we automated most of the operational part of the business. Today, we're not only automating all the operational part, all the analytics that are coming on. The insights that we're getting today are enormous. We're not only impacting cost reduction but revenue improvement. So the impact across the industry is obviously bigger. We're dramatically changing the way we think about engagement, I think, is another way of putting it. We've figured out a lot about how to deal with processes and what we like to say here at Silicon Island, angle Wikibon is, for the first 50 years, it was process certainty, technology uncertainty. And in the last five years or so, we've done a flip where it's process uncertainty and technology certainty. We know it's going to be the cloud. We know there's going to be some set of derivatives going forward, but the difference is, now we need to discover what is actually happening around whether it's the analytics or customers so that we can reveal some of the patterns that will then lead to unlocking new types of value. And data is crucial to that. So is this transition more intense on the data side as opposed to some of the other transitions that might have been more intense on the hardware side or the application development tool side? What do you think? There's clearly big changes on the data side. I would say some of the things that are obvious aren't so obvious. I mean, we're transitioning to the cloud. That's obvious. But we're not transitioning everything to the cloud immediately. So huge transition in where data is located in the cloud, but still massive amounts on the ground. And so cloud to ground integration, how we get the data back and forth becomes a much more critical issue. When we go to the cloud, it's not one cloud. It's multiple clouds. You have multiple SaaS providers. You have some of your own stuff up at Amazon Web Services. So it's not one place. It's multiple places. Now you have to figure out how do you tie the data and manage the data between locations, cloud to cloud, cloud to ground, ground to ground. It's become a much more complicated issue there. Second thing is, it was simple. We all had all our PCs, and all the data was coming in through clients or apps or then web-based browser apps. That was interesting. But now the transition is from internet of things, all these different sources where data is coming in, from click streams. All that data is new and needs to be integrated. We have all that wonderful structured data that we ran the businesses on that were in our big ERP apps, managed that, kept it in our relational databases. About all that unstructured data that's coming in. How do we integrate that? So the challenges are becoming multiplied because of the different sources, different locations of data. So one of the things I want to get your opinion on is what the enablement of the database has done for applications. And you mentioned a few of them back in the 90s and that generation. They became big companies. You could say that IBM owned 100% of the market. Yeah, DB2 is still around. They got some other stuff. But now then you had other database competitors come in. That created public companies. Now we're in an era of the cloud and some of the data challenges you mentioned, these transitions, we've been a lot of us been talking about this application renaissance that's upon us. But where are all the big companies, right? So, or is that just the symptom of the data in that it's native in all apps? So is it, we're trying to figure out that piece of it. What's your take on that? Is there going to be a long tail distribution of apps? Is it going to be a renaissance? Is it just going to be natively changing with data? I mean, you see a lot in memory, so that enables real time. You got batch, you got database. What's that enablement now that you have all kinds of different databases out there? Well, there's a couple of things happening. First of all, the apps world, as we all know, is moving to a SaaS environment and even the biggest player's Oracle is now pushing heavily into the cloud. So there's a transition happening there and there are some new companies being formed. It's maybe a little bit old, but Salesforce is a new company still. It's changed from the Oracle SAP dominance days and Workday and others are coming along, so the apps world has changed. I think that the data itself is becoming more accessible. Now, the word data lake has been kicked around a lot and again, that's reminiscent of 20 years ago when we're introducing data warehouses. It was going to be the panacea, the end all, just throw all your data into the data warehouse, everything will be fine, you can get all your data. Well, people lost millions and millions of dollars doing that and it took a decade before we really learned how to do data warehousing right and understand what was in that data, what to go in, how to process it. We're in the early days of this data lake, which is a new spelling for data warehouse, new stuff in it, but if you don't know what's in your data lake, you're not gonna have success. This gets to this whole question of data management at a higher level. So, databases, if you have one relational database, you have a schema, you know what's in there. Even if you have multiple relational databases, you have a few schemas, you can sort of keep your hands around it. But as you start creating data lakes with all kinds of information going into it, you need a metadata layer. And I'm on the board of Informatica, I'm pretty excited about what they're doing. They're intelligent data lake and what they're doing in terms of managing data across all different types of locations, whether it's on-prem or in the cloud, whether it's in structured or unstructured data, that's gonna give the CDO the kind of capabilities that the CFO has. When you talk to your CFO, you expect the CFO to know exactly where all the money is, right? You ask your CDO, where's all the data? I don't know. Who has accessed all the data? I don't know. Well, that's not a good answer. What's your job? So, a big part of this transition is getting our arms around the governance of data, this super valuable asset. And how do we manage that? What tools are we gonna have in place? Not the individual tools, not within an app, not within a database, but above that. Now, I like to say what Informatica used to do was they would do the plumbing, they would do the stuff between the data warehouse and the OLTP system, and they were the data plumbers. But now they're sort of becoming the heart surgeons where it's still sort of plumbing, but it's much more important plumbing. How do you relate all of that data that you have and keep your arms around it? How do you secure the data, not by building walls around it, but by understanding who has access to it and should they have access? And what's the critical function? I'll guarantee you, most CIO CDOs do not know who's accessing the bulk of their data. So, let's talk about that for a second, Jerry, because historically ownership of some of these IT assets were resident in whoever was paying the bills. So, I owned an application that my team or some group that I was associated with was using. We now have, for the first time, some tools that actually can handle some of that metadata management so that you can reveal something about the data and put some controls on it and take care, have a data inventory for lack of a better term. But do we know enough about how to value that data so that it gets the appropriate level of investment so that we are, in fact, putting the right regime in place to ensure that we're not only generating an optimal value out of it, but ensuring that it remains useful for an extended period of time? What do you think? Do we need to think about data as capital? We absolutely do, and I think we're far from the answer we have the good news is we're getting into a data sharing economy. So, if you think about it, you say apps used to own their data and they were pretty much contained, but data sharing is happening more and we see it here. It's part of the concept of a data lake is let's put, let's all share. We have to be careful how you share, but the value you get. So, another company on the board is Tamer. Tamer's a startup, a mic stone breaker, and what they're doing is they're looking at the long tail of data where they have all sorts of data that can be brought to bear on a problem from all different parts of the company, and they do data curation both with machine intelligence and with human involvement. So they can then look at a Fortune 10 supplier of a manufacturer and all of their suppliers. 300 ERP systems never talk to each other. They have procurement where they're paying one price from a supplier and one part of the business and 10 times that in another part because they were islands. And now when you have this concept of doing the data integration across all those islands, you get great value. The flip side of it is you have to look at the risk. So, there must be management. There has to be a strong metadata layer which gives the CDO, the CIO, whatever you want to call them, the ability to manage that. Great value from integrating but great risk. And it's not just stewardship. It's actually managing it as an asset. Absolutely. So you have also a long history of being an executive within the technology world and starting a number of businesses and now being on the boards of some very interesting companies. How do you think this open source business model is gonna play out because that's so crucial to the evolution of the stack associated with some of these big data technologies? Do we see a clear path for how business models are gonna evolve as people exploit some of these new opportunities and create these new sources of wealth? Well, open source is a very interesting approach to selling software. The first thing you think about open source, oh, it's free. Well, if it's free, really free, then there are no businesses. You don't have a business model, so it can't be free. You may not pay a license fee for it, but if you're gonna really use it in your business, you're gonna pay a support fee or something. If these companies don't have income, they don't exist. So the first thing is to get away from the concept of free. It's just how do you charge for your service? And open source, I think, has a lot of benefits. It fits in certain areas. It brings communities together. It does a lot of good things. But having been involved in lots and lots of companies starting a bunch and being involved in a lot, there's a fundamental. A company to exist must make money. It has to pay its employees. So we have to look at what is that model? Where does the cost come from? It's the same thing as a transition from license to subscription. People like subscription because they only pay a monthly or yearly fee as opposed to this big one-time upfront fee, pop X versus cop cap X. It's great, but it has to come from some place and it has to add up or else you don't have companies. So all of these different varying business models have advantages. They have advantages to the consumer and to the producer of the software. But there's no free lunch. Right. So talk about the dynamics on that business model. You maybe think about Oracle. Oracle has a great business model. I mean, they charge licenses for their databases and they have zillion ways to extract rents out of their install base and they've been doing that for years now. Larry Ellison has made a major pledge and mandated across the company to go to the cloud, which inherently he hasn't said this, but it implies that they have to go to a new license model. So I mean, you can kind of see the writing on the wall. No one wants to renew their licenses at the levels they had. They want to shift it. So your point about shifting how you get paid. So how does the company like Oracle do that? At the same time, you get the small companies trying to compete against the big oracles. How do you see those two dynamics shaking out? The big companies moving to this new transition, specifically Oracle, and then the young startups trying to compete at that big boy table, if you will. Well, I spent a few years at Oracle running the database business, so I know a little bit about it. They have a wonderful business, an absolutely wonderful business. They're very profitable. They're very large. They have a phenomenal go-to-market team and they have a vast array of products. But these transitions are difficult and you have to separate moving from on-prem to cloud and that from license to subscription. So they're two related but not necessarily exactly tied concepts. Having worked with Larry Ellison for a number of years, I'm sure he's gonna figure out the transition. It won't be easy. Now let me talk about something that I do know. I was on the Informatica board for seven years and last year we went private, so all of the board members left and for some reason they asked me to come back as part of the private company with the PE firms that are involved. And it's interesting to see because Informatica was a licensed revenue company and it's now has a substantial subscription component but as a private company, you're able to make some of those changes. There's really a new Informatica being reinvented. It has a phenomenal set of products, a lot of innovation, all their products in the upper right hand corner of Gartner. But as a private company, they have the capability to make some of these changes sort of under the covers without Wall Street banging on the door all the time. So- Hey, explain how significant that is because the board member, you've seen, you've been on both sides of the table, public company, board member and private. I mean, Michael Dell took us coming private, buys EMC. I mean, people in the top executive ranks are saying private is the way to go because they can do creative things, they can actually do the long game. Explain the importance of that and the move by Informatica. So there's no simple answer. There's absolute benefits of being a public company. There's a lot of capabilities you have, a lot of sources of capital and so on. There are definitely advantages of being a private company, whether you're a private startup, trying to become a big company or whether you're a big company who's gone private, both of those, so basically Informatica is now a pre-IPO company, right? Because it's a private company and eventually, so it's just like Cloudera or whatever, it's a pre-IPO company. A lot of incentives that you can put in place, a lot of things that you can do to restructure the company, but eventually you probably go back out in your public company again for the benefits of access to capital and so on. Trade-offs, always trade-offs. But the scrutiny of the 90 day shot clock as they say is can be can brutal. It's absolutely can be. And especially when you match it with these transitions, when you have the transitions in industry and Oracle will have its issues as a public company making that transition. I have confidence that they'll figure it out. But it's much more difficult in the bright light of day. Jerry, final question, because we're getting running tight on time now in the segment is you're on a lot of boards, which means you see a lot of management teams, you see a lot of market forces, just generically anecdotally, can you share the pattern that you see in this transitional period that we're in this industry? What are the common patterns that you're seeing across not only the board you're on, but also other companies that you see on the marketplace? Is there anything that pops out that's trending that you can see? These are the top three things I continuously hear in board meetings and from customers and the customer's customers. In terms of the management of the? No, just in terms of the marketplace, what the, is it informatic as customers are looking for and their customers are hearing? What are the biggest items that are on the business boardroom table that are being discussed? Now, in all the companies I'm involved with that are focused on data, data is becoming much more of a boardroom topic in our customers. So, data not only to make the company more efficient, but to generate revenues, to be closer to customers. Another one, real-time mem SQL is another company I'm involved with. Absolutely a big, big transition where minute by minute, you have to be able to guide the drill bit. You have to be able to make offers to customers. You have to be able to schedule your Uber. All of those are having much more direct impact on how you run your business minute to minute as opposed to how you looked at your business. So it's sort of proactive as opposed to reactive. Final question, final, final question for the folks watching out there who aren't here in Big Data Week in Strata Hadoop. What's happening here on the ground this week? If you had to summarize, kind of globalize the trend here at the event, what's the big trend that's that you can say, this is what's happening here at the moment? Well, what's happening is a 10th anniversary of Hadoop and already MapReduce, one of the components is sort of history now and Spark's taking its place. So there are transitions already in only 10-year-old technology. But what you can see as you walk the floor is the impact that data is having in business. Every booth you go to, no matter what company, they're telling stories about how data is impacting business and it's not in the back office anymore, it's in the front office. And it's not yesterday and looking backwards, it's looking forward. So I think data is playing a much, much more important role. We as an industry have to get our hands around it. We're not done yet. This is only, this is still only the beginning. Over most of it. And this for 40 years, but we're still at the beginning. It's still exciting and there's a lot more to come. It's super exciting. Just like the Cube, we're on the front lines. We have all the action here inside the Cube here, live in Silicon Valley. Of course, we follow us. We're gonna Hadoop Summit in Dublin next month. We're gonna also be there. Jerry Held, thank you for sharing your insights on the Cube. Really appreciate it. Thank you for your time. Mr. Cube, we'll be back with more live coverage here in Silicon Valley in San Jose after this short break.