 Hey, Jeff Frick here with The Key. We are on the ground at the Cosmopolitan Hotel in lovely Las Vegas at the Predix. First Predix developer show, 2016, 1700 people for the first developer show. I still find it hard to believe. I'm joined by the next guest, Tia Hellman, the chief data officer for GE Oil and Gas. Welcome. Thank you. So exciting times. You got 1700 people here in all the sessions learning about how they can develop in the Predix platform. Who would have thought three years ago that we'd be here today? Yeah, it's all about scale, right? It's unbelievable how it turns out to be an event that actually is not just 1700 people, but 1700 people from all over the world. Right, and developers, and everyone is competing for the developer's hearts and minds. Everybody wants developers because everybody's got an ecosystem and a platform that they're trying to get people to develop too. So what is it that you think is attracting such a great turnout here for your first event? I think it's a Predix overall, it's a pretty unique entity for us. And people realize it now, it's not just for GE, but it's truly for the world, something that has unique capabilities and features when you think about the industrial internet that nobody else has to offer. And that's where you see people getting interested and trying it out, kicking the tires, and ultimately now competing, building applications on top of Predix. Right, and you're at Oil and Gas and Oil and Gas is just this funny industry with such high variability on the price, right? The price of a barrel of oil changes from not that long ago, it was over $100, then it dropped down to 26, I don't know what amounts are back to like $45, that was over 50. So from a manufacturing point of view, and these guys are building facilities and production to try to figure out how to make money at a market rate that they don't have a ton of input on, they got to look for every ounce of efficiency that they can, and actually be constantly changing the drivers for how they run their actual manufacturing and operations. That's exactly right, because volatility is the name of the game in Oil and Gas. Not just that the oil price has dropped from well over 100 down to 26, the beginning of the year, also the recoveries from 26 back to 45 are pretty quick. So for our customers, for our industry overall, one of the key drivers is to be flexible, nimble, very fast in the way you ramp down and ramp up your capacity, because the worst thing that can happen is you'll slim down, you'll tighten the belt, but then once the recovery comes, is you're not ready to execute. Yeah, that's interesting. So they're really dynamically tuning their outputs based on real-time economic factors. It has nothing to do with kind of what they could get out of that hole in the ground if they really wanted to. It's the output that counts, but think about it more about the investment levels, and the investments in the CapEx and operating expenses. In times where oil price is seeing volatility like this is you need to look for new business models and operating models that allow you to be extremely flexible in the way you operate so you stay ultimately afloat and be a profitable business over a longer term. And that's one of the key things that we think about. Our customer base needs to be able to operate effectively and efficiently and profitably at $30 a barrel, as well as $45, $50, or even $80 or $90. And in times like this where, and let's be clear, the oil and gas industry is an asset-intense industry. I mean, there's lots of equipment, costs a lot of money. So in times like this, you need to be smart about your investment levels and where you deploy resources and how do you make the most out of the money you invest. Right. The other thing is you got to go where the oil is and you got to try new places. So in terms of kind of difficulty of climactic conditions on that gear, which then has to be connected to computer gear and software, totally unique challenge than in a pristine office or a climate-controlled data center where you've got all these devices in this edge. So tell us about some of these edge-specific challenges because you've got to get compute there, you've got to get connectivity there, and you got to get storage there. And then you got to figure out how much of it's there at the bottom of the deep sea well versus in Texas or back at the data center and how do they move that compute and store and decide which goes where. Right. So look at, I mean, GE is always looking for technological challenges. So that's now a sweet spot. You mentioned subsea sensors. How do we understand what's happening in subsea reservoirs? What kind of sensors do we need to be able to operate continuously in a harsh environment? Operate also at an operating cost that is ultimately working with the oil price environment you're in, and these are the items in the hardware that we care about, especially in my business, we care heavily about sensors and condition monitoring capabilities which are behind your back here. Right, right. And you can establish an operating rhythm and transparency and visibility in items that you normally don't see. Right. And that's where you need sensors and measures and then condition monitoring that allow you to then ultimately make the right decisions about the operating configuration of the production environment. And I think a lot of people kind of understand the sensor concept, but they don't understand how many other kind of layers there are in that communication chain between the sensors and the controllers and the fabric and the network and the individual device and then kind of the system device and then the field. There's a whole bunch of ways that you can slice and disorganize this output in this organization, really the orchestration of this system. That's right. Well, we call it the on-ramp. You know, the on-ramp and one of the key drivers for the on-ramp is obviously is to be able to scale this. In today's environment, you're looking at replenishment times of one second intervals, right? With thousands of tags being connected to the internet and then you have to make it cyber secure. And nobody wants to know. Small little detail. Small little detail. Past that pesky security thing. When you collect reservoir information, that's a core DNA of every owning gas producer in the world. That needs to be something that they want to protect by all means. And we need to have technology. We have to have the means to provide them this confidence level that their data is secure and that they can ultimately utilize that data in an environment that they have complete privacy around them. Right. So just love to get some of your insight, stories from the field, if you will. I'm sure you're out talking to customers all the time. When they start to integrate and put some of this stuff together, what are some of the things that are coming back? Like, wow, I didn't know this could happen. Or wow, I never really thought of it that way. Or wow, now I can get beyond worrying about squeezing out a little bit more operational efficiency, which is good and worth a lot of money. But now I can start, you know, I can change the lens in the way that I run my business. Right. Well, we heard this this morning in the keynote session is there's a very profound difference when it comes to the industrial internet is the amount of data, once you turn these assets, these machines on, that you're capturing is just tremendous. I mean, it's horrendous is how much data comes in, right? Now, the second piece to it is the equipment was designed to operate efficiently. And it's me, you're looking basically for a piece of equipment that on the normal circumstances, one's perfect. And so you're looking at a piece of equipment, you're collecting all this data and you don't really find something for it. It keeps on running. So there's nothing to really look for. So you're looking literally for the needle and the haystack, right? But this piece of information, ultimately, if you get it, can then drive this massive productivity improvement. So industrial internet, massive amount of data. In oil and gas, a further complication is, as you said, is we're not in the shiny office in Las Vegas, downtown San Francisco or Chicago, Boston, is we're in kind of the rough place of the world where you got to think about connectivity, whether you're in Africa, whether you're in Latin America, even if you're in the North Sea, is these are areas where you need to think about connectivity and that's where we establish with our partners connectivity environments to be able to service our customers, right? And then second, around this, as I said, is too much data can also be difficult to digest. So what are the compression methodologies that you need to deploy in order to ensure that you really look at the really relevant data, not just everything that is creating just noise, right? And we have done this. I mean, if you look at the equipment that in the oil and gas industry is deployed by GE, especially around turbo machinery, our turbines, you know, they're creating quite a bit of data. I mean, just to give you a Wal-Park reference, is our equipment in the field generates as much data as Facebook does an all day. We got 1.6 billion users, we've got about 15,000 turbines out there. Just to give you more- And the day-to-day, the day-to-day comparison, right? The day-to-day comparison, right? It's pretty close, so to say. Yeah, it's basically the same kind of data creation. Right, right. Now, not all the data is useful. Now, on Facebook, all the pictures are important. They're stored everywhere. Every single one. Now, we have to make sure that all the data that we're collecting is just filtering out the important piece. Right. And our data scientists, our engineers think about compression methodologies, filtering methodologies to establish kind of the right data and filtering out the noise so that actually we can focus on the real things. Right, right. Because one of the things that we see with operators in the field, when everything is blinking, you don't see the force because you're too close to the trees. Right, right. So, give you the last word before we're out of time for people that didn't make it here to transform. Share a little bit of what's the vibe. It's the first one, some of the hallway chatter and some of the conversations that you're having to share to people that they should probably come next year. Well, I think it's amazing that we have lots of developers here who truly think in business use cases or customer use cases. So it is truly amazing to see that it's not really kind of a little kind of a laboratory experience anymore. People come here from customers, from partners, always eye partners with true use cases in the oil and gas industry and where they're trying to get their hands on. How can I use predicts in the operating system capabilities that G is offering in the real world? And I can say for the oil and gas business that we probably have more use cases where we have people to staff against them right now. Which is a good problem to have. It's a good problem to have. So, Mia, if we have 1,700 people this year, you probably should be hitting like 3,500 next year. Just double it. I don't doubt it. I mean, the rooms are packed. I went in a couple of session rooms earlier today and I was in the one with the digital twin and people were taking pictures of the slides and paying close attention. Yeah, and then they mean the dojo one is interesting where we're bringing customers and partners in and they're sitting for a week in a room and basically building a real application. So no more just kicking the tires doing the real thing, actually the real test life. Right, and that's the test of a platform as other people are building applications on because nobody's got a line item budget for a new platform, right? It doesn't exist. It's a pretty, we went through our bumps, we had our speed bumps along the way. I think as we are just timing wise and the scale we have now in the experience, the lessons learned. We're certainly ahead of the rest of the pack and bringing partners like Microsoft on board allows us to then bring other folks on board because they say is what the heck why should I invest in a platform if I can build the application that actually drives productivity. Right, all right. Well, Matias, thanks for stopping by. Appreciate it. And... It's a real pleasure. It's real oil and gas, it's a real industry right here. This is big, heavy stuff. He's got a little demo over here and all the boys are hanging out and girls playing with the toys, which is what we all like to do. So thanks again. Appreciate it. All right, Jeff Frick here with theCUBE. We're at Predix Transform at the Cosmo in Las Vegas. Thanks for watching.