 Hey, welcome back everybody. Jeff Frick here with theCUBE. We're on the ground at the Cosmopolitan Resort in downtown Las Vegas. It's 120 degrees outside, perfect for a conversation about IoT and industrial internet where you can't be in a beautifully air-conditioned taking care of data center all the time. So we're really excited to be joined by our next guest who's been playing in this field for a long time. It's Hema, we call him up. He likes to just go by Hema, likes Bono and Brittany and share all the famous people. So welcome to theCUBE. Thank you, thank you. Interesting conversation, you said you just come off a lot of partner conversations here. We were commenting for a company doing their first developer conference that 1700 people show up. Pretty impressive feat and obviously you need a partner ecosystem. So what were some of the conversations with the partners you were having? Absolutely, and I've been in the valley for long enough and I attended the first Java one and it was as small as this. And that's what I'm thinking of where we'll be in 10 years, right? I think one of the key things about a platform is that's what Harrell talked about in the morning. We don't do all of it, right? It's an ecosystem play, partners have to add to it. And so the conversations I was having was with a group of partners who had APIs on top of the platform. And so one example is a good need in the cloud space is monetization and metering, right? If you think about if I'm a service provider, how do I know who's using what? How do I segment the customers and give them different price points and sort of dynamic pricing? Right. And so one of the partners I talked to was give them an API. You as a service provider or an application provider, just call their API, go to their portal, look at the analytics on who's using what and change the pricing, right? That's one example. The second example partner I talked to was anomaly detection. A lot of the industrial use cases are about detecting anomalies in the data. You don't want to write that from scratch as a developer, right? So this partner gives you an API, you push the data, they sort of do the anomaly detection, they give you the outcomes. And so we're trying to have this conversation about how do developers get access to the APIs so that they can take down the development time from months to weeks to days. Right. And the other play you've been around for a long time is nobody buys a platform, right? They buy applications to solve business problems. So if you don't have an application store, you can never kind of get to the platform, Nirvana, that you're trying to get to. And nobody can do it all. So as we see in all the shows we go to, we go to a lot of shows. Ecosystem is really, really important. No one company can do it all. Absolutely. And I know I've been on the other side of the spectrum where we tried to do a platform, it didn't work, right? Because it's a hard sell. What's actually very different with R-Spaces, it's much harder. Because if you go talk to a VP of operations or any CEO of an industrial company, they care about two outcomes. How do I reduce my unplanned downtime and increase productivity? They don't care what's the platform underneath that, right? Right, right, right. So it's always- No line item budget for platform. Yeah, exactly. It's all about am I managing my assets better, am I monitoring them better, and am I leveraging them to make more revenue, right? Make more power, make the trains go longer. And so to make that happen, we cannot scale as fast as the market needs us to, right? Because the developers are trying to build these applications at a pace that is unprecedented. You know, Scott from Microsoft, he made this statement in the keynote today. It is the coolest job today to be a developer, right? And in this context, you have so many developers just trying to do a lot of cool things. And so the partner ecosystem just accelerates that, you know, paths to market faster, yeah. The tough challenge for you guys though, which you're obviously having some success against, is everybody's competing for the developer. Microsoft's competing for the developer. You know, Amazon's competing for the developer. Everybody wants the developer. And I just, I love kind of the GE commercials now that they run on television with the software guy that's trying to convince his friends he's not working on low-to-moke, he's actually working on software. So what is, you know, how are you selling that into developers? What's really the thing that's hooking them to get so many of them here for this inaugural event? I know, I have a funny story to tell. I went to Stanford Engineering School for recruitment the first year I joined. I was there for an hour, not a single developer show. How long ago was this? Four years. Okay. We did the same earlier this year. You won't believe how much interest there is. I think what's happened is, the sort of the trend of IoT has become, the market has become more aware, but then people understand industrial IoT is a problem by itself, like it's a different scale. So what we are telling the customer, developers, it's a cool thing to do. You know, as technologists in the valley, what do you care about? Is it a cool technology that I'm going to develop and do I make a real outcome? Right. With this, with predicts, we're able to tell the developers you do both. Cutting edge of technology. We're talking about blockchain. We're talking about microservices. So we're talking about technologies that are there out there, you know, ahead of the market. And the second one is, the key to success in the industrial space is customer interaction. Because I walked into a locomotive re-manufacturing facility thinking I could do a mobile app with touchscreen. I found out they're wearing gloves. Touchscreens don't work. Right, right. And then they're looking at virtual reality kind of a thing. And so the actual customer sort of feedback and how they use it is pushing us to do a lot of cool stuff. And so that's what one part of telling the developer is, this is real. You got a lot of new cool APIs. It accelerates the development you need to do. And then the second part of it is, you know, the outcomes, the business outcomes. Being a technologist in the valley, having built platforms, I've always been intrigued by what is the real customer outcome. Right, right. With Predix, you have so much closer to the customer outcome, because you're sort of merging the physics and the math and the math and delivering outcomes closer to the customer. Yeah, yeah. And as you said, you're working with all the same tools. You're working with open source. We've been here for a couple of hours. We're already talking about Docker. We're talking about microservices. We're talking about all the stuff that you hear about at every other conference, whether it's EMC World or VM World or Oracle Open World, or it's the same types of technologies really bringing to bear. But it's funny, your perspective, you've been around for a while, as you've got more exposed to the operations technology side of the house, right? And this integration of the two, what's kind of your thought as you've kind of brought your IT experience into this existing world? Like I said, they're wearing gloves, right? It's harsh environments. It's not all super controlled. As an engineer, by degree, by profession, we like to touch things, right? In code, it's very virtual, right? What I've loved in the first year I was here, I visited an airplane manufacturing facility, right? And I saw how it was being built, how software can make a change. So for me, switching on to the operations side gave me a real perspective of what things are really used for. My previous life was building banking software, there's nothing to touch in a bank, I'm still here, right? And so, I think that's what's been interesting for me and when we talked about Owen, that's exactly how my team felt when we started here. Everyone would ask us, what do you do at GV, you know? They don't do software, they build airplane engines, right? But once we start telling them the story about how the kind of outcomes, right? Make a plane go faster, train go faster, make more power, these outcomes tend to rhyme with developers because I think software has come to a place where, you know, I use this metric, 80% of the software that gets written doesn't get used, right? In this space, the customer don't let us do that because they're driven by bottom line and top line. And when they're driven by that so much, you have to show them real value and then you only do work that adds value. And so that's been the interesting transition for me from this pure software world to a merged space where you get to see real outcomes. Right, you get a touch, it's metal, right? It's big, big machine. Everybody likes big machines, right? From the time you're a little three-year-old, everybody likes the big truck that goes by. And it's easier for me to tell my daughter what I do now. Like, what do you do? Oh, you're flying on a vacation, I make that plane run go faster. Yeah, yeah, that's great. Well, let's shift gears a little bit again on such great connectivity between IT and OT in your past life. You've got a patent on, it's a very long title. People can go check it out on LinkedIn. But the part that jumped out to me was the occasionally connected devices. And we think about all these value chains and now these software chains. And it's an API world and it's an app world and it's connected to the cloud and it's all this stuff. Well, sometimes the connections just don't stay. But stuff's gotta still happen, something's gotta happen, and things gotta continue to work. So have you kind of taken that experience and applied that in this OT space? So it's interesting you mentioned that, right? When we started Freedix, people thought we were a cloud platform. We are not. We are what we call as an edge to cloud platform. Edge is all about disconnected world. If you go back to even two years back, the operational side was completely air gapped. There was no connectivity between that operational side and the cloud side. Now, when we try to open it up, right? It doesn't mean they're always connected because the operating environment's deep sea oil rig, you only have satellite connection at very exorbitant prices for a short time. And so you may have connection, but you don't wanna burn your dollars with all the connection. So occasionally, connected sort of applies into this industrial world a lot as much as mobile because here, price is premium, connectivity is premium. And so what we are starting to do is give them the capabilities to build these applications where they can assume that there's no connectivity, store the data locally, run analytics locally. And so what we talked about in the keynote this morning is this notion of a digital twin. How do you take that digital twin and move that to the edge? And when it's in the edge, it is not connected to the cloud. It may be connected occasionally, but when you're completely disconnected, how do you still use all the analytics? How do you still do all the outcomes? And I'll give you a good example. In the past world, when you want to make more power, you would have to send the data to the cloud or somewhere in the local M&D center and run an analytic that says make more power or not, right? When you make more power, the life digweeds because you got to operate at a higher temperature. We push that analytic with all the data to the edge. Now you can make the decision on, do I optimize on the life thing or the power production based on how much money you're going to make? So this is this notion of disconnected nature at the edge so that you're making decisions in real time where the data is and where the assets are. So I come from a mobile space. We thought of disconnected in the mobile space. It was a sort of similar transition into industrial space. You know, the technologies are almost similar. We have a database that runs on the edge. We have a database that runs in the cloud. We have to occasionally sync the data. The kinds of data are different. The volume is different, but it's similar model, right? That's interesting. You're the first one that kind of tied it back to a mobile which is such an easy analogy because we all carry it in our pocket, right? Sometimes it's connected, sometimes it's not. Sometimes you work independently, sometimes you're on the web. At some point in time they get re-synced up again. Exactly. So when someone's kind of starting this journey, say it's an old line company, they've got good systems, they work, everything's working. Where do they start? Do they start with the edge device? You put a little bit of compute and store next to things? Is it, you start with the low hanging fruit on what you can send to the actual cloud? How do they kind of sort through the decision tree to figure out what goes where, when? And so that's an interesting question because we see customers in every different segment, like some of them are very comfortable with data moving to the cloud, right? Healthcare has done that, transportation has done that. There are some businesses which are heavily regulated, which they want to keep the data in the operational environment. So it's sort of driven by the use case and their comfort of how far can they go with sort of going beyond what they've done, right? The easiest approach what we tell people is, if you already are collecting data in the cloud, and I say in the cloud, it's sort of a not real cloud, but today we have MND centers, monitoring and diagnostic centers for our wind turbines. So that data is already being sent to some central location. In that case, what we tell them is, extend the data, write analytics on that data so that you're delivering this predictive maintenance capabilities. So that's an easy transition to do pure cloud. In some cases where they don't have any edge sort of data collection, we ship with and we demoed this in the keynote today, we have a box that they can put right next to their asset, could be a turbine, could be a jet engine, and then collect the data and then do some local analytics, and you have nothing to do with the cloud. And then you just manage all of these assets, right? So it's been a combination of which is the customer, what domain they are in, how regulated they are, and sort of then decide based on the use case. The other thing we tell them is don't go big, right? It's just hard coming from GE, right? We think everything from GE is big. I know. And so telling the industrial guys, absolutely don't go big. And it's good that Jeff and our leadership was thinking of fast works and lean and sort of pushing that message within the organization. Now we are sort of leveraging the same thing and saying, hey, this is software. We are a lot more agile than any of what you've seen. So let's go incremental gains. Let's find out the smaller value. So start with, like you said, small edge or small cloud, and then grow that as you show more outcomes. Yeah, it's great because everybody we've talked to has a big software background at GE. It's amazing that they've done such a good job, Bill and team bringing in really software DNA. And that said, you're in big industrial heavy stuff that's been running a long time, but you're still kind of pushing the edge. And I want to bring up a topic we talked about. When you first walked up, and that's blockchain. A lot of stories on Bitcoin, everybody kind of knows the story of Bitcoin, whatever, I was thinking Bitcoin is an application, blockchain is a platform, but we're really seeing a lot of investment of effort and at least conversations on the blockchain front from IBM, SAP and you guys, what's your kind of take on the potential for blockchain, kind of the timing for blockchain? How do you see, where does it fit and what's the timing on how it's gonna slowly increase its importance in the marketplace? I mean, and you and I talked about this before, it just caught fire like crazy in the last three months. You know, the conferences have gone from 20 people to a thousand people in less than like three months. What's interesting is this whole transformation that the industrial economy is going through, the companies are going through, it's enabling us to sort of insert these technologies which by themselves are hard to digest, but now we are telling them you're going distributed, right? Critics is a distributed architecture. As part of that, let's think about a distributed ledger so that we were talking about an aviation use case. When a plane comes into a service, the service is actually done by at least 20 different companies, very manual, and there's one centralized sort of reconciliation of all of that. Now we're telling them, ignore the centralized reconciliation, run a distributed blockchain, and then each of you gets to know all the transactions that have been done in the blockchain and non-reperiate it so that you all can be verified on what you've done. They're like, hmm, that's interesting. I've never thought about it. How does it reduce my apex? How does it reduce my capex? And so it's triggering those thoughts because it's no longer about, oh my God, that's a new technology. We have crossed that bridge with the cloud sort of transition. And so that's one example of where we are seeing the adoption of blockchain. The other interesting thing that's much closer is the digital twin. The whole point of digital twin is you have a virtual model of the asset and all the configuration changes it's going through are captured somewhere. But what's key in the service industry like ours, it has to be attested by someone. Applying blockchain when the servicing of that particular asset is done by multiple players and each of them is being verified is awesome. That way, then the technology sort of a curve is lower in terms of how much investment you have to make. And then there's complete non-repedation on that. So I think from just a timeline standpoint, I think the next six months are going to be like really strong POCs. And people might call it early adoption curve, but the next six months is where I see, proving out that this applies mid to early, mid to late 2017 is when I think we'll see the larger adoptions and real projects. You made a really interesting comment about once everybody kind of got over the hump with accepting cloud, now suddenly they're pretty open to trying new technologies. My buddy, Gary Ornstein from MIMP's sequel tweeted to me during the keynote, Jeff, what databases are they talking about? You're at Sybase. I'm like, Gary, they're talking about every database. They're really kind of horses for courses depending on the application. So the ability to adopt, integrate, try lots of new technology seems to be really taken off within the industrial space. I think people think it's a lot of, the transformation is a lot of technical, but it's actually a lot cultural, right? To make them think about change and technology and cloud, how do you make them think as a culture change is acceptable, right? Because they're so resistant to change because if you go look at FD airlines and how much verification they have to do, they're not acceptable of change. And so it's a part, half technical, half cultural, and you get them to the cultural change, accepting the cultural change, technology is like, it's easy, it's not as hard as it used to be. Right, right, that's funny because we say it all the time, it's people processing tech and the tech's actually the easy, the people in process is the hard part. And what we're finding out with the cloud transition is the people, right? Like typical traditional IT, they may not have a role now. If it's all cloud, what are they managing? Are they managing systems? No, because the cloud manages it. Now everyone's thinking like, I'm gonna write code, I'm a developer, that's the transition we are going through. Right, he might give you the last word, running out of time, top two priorities for the next six months. Digital twin and edge. And then I'm going to add a third one because we mentioned it in the keynote, service availability. So we call it service quality, service availability, the two trust for a cloud is the uptimes. Those are three main areas. We'll keep an eye on that and we'll check in with you again at Minds of Machines later this year. All right, thanks for stopping by. Bye, thank you. All right, you see, Mom, Jeff, you were watching theCUBE. We are in Las Vegas at the lovely Cosmopolitan Hotel. Thanks for watching. Thanks.