Ep. 184 Wesley Powell, Chief Executive Officer at Open Origin | Data Center Go-to-Market Podcast

We fully insure up to tens of billions of dollars. The full cost of the full replace cost of the data center with our technology reinsurance program. And what that means is that if something fails through our thing, let's say our superconductors fail because people consider that risky even though it's 30 years old. If that were not to work or we had to change the vendor, the insurance company will come in, pay for the entire replacements, cover all the liquidated damages to the customer while that's not functioning. So lost revenue, business interruption insurance, the whole nine yards. Fix it. Right. And then bring in a new vendor and repair all the stuff that needs to be done and get it working back on spec again, all on the insurance company's nickel. Right. So we may not have $3 billion to come in to completely replace the microgrid on a $10 billion data center, but the insurance company does, and they're willing to stand behind it with that level of insurance because we went and worked with them to get that available. So we've created this new stack of extended technology risk insurance that simply is not available publicly on the market. Hi, this is Wesley Powell with OpenOrigin, and you're watching the Data Center Go to Market podcast. Hey there. Joshua Feinberg, host of the Data Center Go to Market podcast. And today we a very special guest. Joining me, I'm welcoming Wesley Powell, who is the CEO and founder of Open Origin. Wesley's based in the New York City area. Company does a lot of business in Texas and throughout the United States. Wesley, welcome to the podcast. Thank you so much for having me. Oh, it's my pleasure to have you here as well. So the big question that we're going to be talking about today is what happens when infrastructure, energy demand, AI demand, collides with outdated assumptions about energy infrastructure? The biggest bottleneck for AI growth, AI demand, AI infrastructure that we've been talking about for 12, 24 months at this point is no longer just the compute. It's the energy availability, what's available on the grid and energy deployment speed and all the outdated assumptions about centralized electric supply. When you think about all of this, what do you feel is the biggest go to market challenge for a company like Open Origin? Well, as you're probably aware, we have been talking about this for a long time now. Right? So we're about 24 months into that to the hype cycle, let's call it, of the identification of the problem and then people starting to try to address it. So right now, I would say the biggest problem is cutting through the noise. We're in a situation where everybody and their brother is trying to sell power to a hyperscaler data center because the amount of money being spent there is in excess of any amount of money ever spent on any particular commodity ever, right? So it's just a huge target. And as a result, you've got folks that are in real estate, you've got folks that are in traditional, you know, PPA power production type developers who have energy facilities which aren't really a match for data centers. But hey, any port in the storm, right? And you, there's just so much noise that the buyers are having to evaluate 10 times as many opportunities as they were, let's say two years ago or maybe even 20 times more as a result. It's really crowded and the biggest challenge, I would say is cutting through the noise. When you talk with your existing clients, your existing partners, you feel that people are ready for all of this, that they're really emotionally prepared for what a massive AI infrastructure build out looks like. On a national level, I would say no because a, nobody's actually done it yet, right? The folks who are doing it now are still in the process of building the first generation of the truly massive gigawatt plus facilities. They're just now coming online, having been started a couple of years ago and to my knowledge yet I'm not sure that a true 1 gigawatt total power facility like in one data hall has gone live yet. I don't believe that's yet the case. There are several that are under construction, but they're still moving in that direction. So I would say that no, they, they are not emotionally prepared for it because no one is. It's, it's, it's inconceivable that people don't know what a gigawatt is. It's just so much electricity. We don't really have a easily identifiable amount of, of usage, shall we say to demand, to compare it to. That is itself not a silly number. We start talking about millions of households, right? Well, I don't know, like it's hard for me to comprehend even that number. And I'm in the business, right? How much power does a million households actually know? I mean like a million is a big number. So I think that's really where everyone's still struggling a little bit is that we went very quickly from 100 megawatts to gigawatt to multiple gigawatts in a period of, gosh, 18 months. Nothing else in human history has scaled like that. A Lot of people talk about the concept of behind the meter self generating off the grid. Which parts of those are open origin specifically working on solving. We only do non grid connected islanded. And the reason why is because the grid was not designed for this. The grid was designed for to distribute a little bit of power to a very large number of users. It was never designed to deliver a lot of power to a single big industrial user. Historically those were all managed with dedicated power plants that had power going directly to that side. In fact in some cases they were dedicated to a single machine. I grew up in southeast Kansas and little personal anecdote. There's this giant electric, effectively the same as a steam shovel, but all electric called Big Brutus which was the largest one built in the United States I think. And it was like 10 miles from my house. And they had a, they had a cable that went like 200 miles that was the size of a pickup truck. The cable, it's just solid copper that powered this thing as it dug its way across southeast Kansas unearthing coal back when coal was a big shipping thing in the early 1900s, up through the 1950s. And it had a dedicated power plant. I mean that's how you did it back then. You built power plant just for the thing. And the grid was not involved in those historical conversations around this. For steel plants and other big loads, Data centers are 100 times the load of the largest aluminum and steel plants. You know those, those were the only things that were even close to gigawatt scale. In fact, the largest, you know, aluminum plant in the world in is in New Zealand and I think it's 12 gigawatts and it's got a dedicated hydroelectric dam. Right, that's just for that. And so these are the types of things that were done historically around those, those types of loads. That's what we need to do today. Right, except that now we need to figure out how to do that in West Texas or New Mexico where we don't have a giant lake that we can tap into for those gigawatt scale power plants. And so we need to do something else. And slapping together a bunch of, you know, small generators is probably not ideal for many reasons. Efficiency and cost not being too very top end ones, not to mention the pollution factor. But we've solved that problem really with solar and battery. We've spent about 18 months developing a very customized solution that is able to provide Tier 4 power completely from a renewable perspective while using a very reasonable amount of land. So we're much more dense than your typical solar array. In fact, we're right about an acre per megawatt. So that's a huge advantage compared to historical 3 or 4 acres per megawatt deployments. So we've really reduced the land footprint and as a result we're much more efficient than that and we're cheaper on the whole. We can build the power plant and the whole dedicated power facility, including the data center and all the redundancies, including the backup thermal generation for emergency power, at about a half to a third of the cost of a traditional grid connected data center. And that's really phenomenal. So it's fueling all of this. The underlying root cause of all this demand seems to be the hockey stick growth of generative AI starting in like late 2022, early 2023, if we go back five or 10 years before that, people were already fairly used to using AI apps on their phone and streaming and maps and all those other that were being influenced by AI. But it wasn't until really Sam Altman and OpenAI and Jensen Huang and Nvidia became household names that we've seen this explosion of growth. So given that AI is really fundamentally changing these power assumptions, how have you seen these high density AI workloads impacting how you go about infrastructure planning? Well, that's very interesting because we got into this business in 2018, so we actually founded Open Origin to solve this problem. I ran an AI company back in 2018 and, and we ran into the problem and I couldn't find a host. I was running around looking for somebody who had a very small amount of power, like 20 megawatts or something, and looking to find the right density to get to where we needed to be. And this was Pascal generation stuff, this was 2018, 2019 Volta level stuff. But even then we were looking at rack densities that needed to be at about 100 kilowatts per rack and nobody had it. It just was not available. I could get 50 from Equinix and those guys, but when I started, you know, putting a, for what they were charging, it was not cost effective at, you know, to half fill the racks, right? And I really was like, okay, this is, this is actually the problem that needs to be solved. It's bigger than the AI problem. So I got out of the AI business and I got into the AI energy business and we, we, we spent the next four years developing the solution for that and brought it to officially to market in 2024 for us. Right? Like we were going out there and we were going to build a plant just for our, for our own use. And by the time that we got it designed right as we were finishing that, the whole AI blew up and became this huge energy thing. And we said, hey, you know what, we should just pivot to doing that for other people because this is, we're four years ahead of everybody else. Turns out we were actually a lot farther than that. We were probably more like six or seven years ahead of everyone. And in fact, our solution that got presented at Cirid last summer in Europe and was held up as the way to deploy power in the future for data center and off grid deployments. Because we combined a multitude of energy efficiency and enhancement designs into our solar powered data center power plants that we call the UPP or the uninterruptible power plant as a UPS play on words. And we had things like superconductors moving the power around. We only, we only did dc. We took DC right off the solar, dumped it right into the, into the plants, into the batteries. And the batteries power the DC all day. That's kind of things. These were really, really innovative uses of actually widely used tech, but no one had ever applied them that way before. So for example, we don't have any transformers in our design. We only use power electronics when and if we do need to do a transform, a transformation going from DC to AC. And so it's very, very efficient. It's in 99, you know, 99 and a half plus percent efficiencies and we only use low voltage. So we go back and forth between 1500 v DC and you know, 483 phase on the, on the AC side. This keeps the cost way down for the microgrid and allows us to move power around the plant almost completely losslessly because the superconductor is of course 100% loss free while it's on the superconducting side. So we're, this was, this was huge. So we're very far ahead of everyone from the design and deployment perspective. And that also makes it a lot cheaper. It doesn't hurt that the price of solar and batteries has come down 90% in the last five years either. So, you know, it so happened that the market is now aligning to where we are, which I think is quite interesting because we saw that five years ago, we brought it out two years ago. We were a little ahead of the marketplace from an understanding perspective. And Only now in 2026 are the big hyperscalers coming to us and seriously considering what we've got to Offer. What's really interesting is just like you, I was evangelizing AI in SAS in the late 2000 and tens, but it was only last fall, maybe about eight or ten months ago, that I heard the word superconductor for the first time. And now you're the third person that I'm talking to about superconductors. Because in all the time I'd been working with colo operators, starting in the early 2010s presence, like no one had the need to talk about what a superconductor was. Well, yeah, let me, let me tell you something really interesting. The crazy part is that internal DC superconducting cabling has been available as a product since the late 1990s using liquid nitrogen cooling. And I went back and dug up those old websites and found out who those suppliers were. And I went back to the people who bought those suppliers and we worked with them and over a period of 18 months developed a technology that not only allows us to move gigawatts of power around outside on the giant cables, but, but we actually designed the data center so that we can support out of the box, our smallest power drop can take 15 megawatts per drop. And so we actually sell that to the customer at 1 megawatt per drop just so that. Because otherwise they freak out. But we provide that and the cooling. So when we build you a data center, you don't just get the power drop, but you also get the cooling as well for the same price. And here's the kicker. We don't charge for the energy. So you're paying 145 your kilowatt month, just like you would for any other good tier 2 plus data center. But you're getting tier 4 reliability out of one of our centers and you're getting the power included as part of that price. With all of these points and considerations that you're bringing up that probably 80, 90% of the people watching this hadn't truly thought about quite yet. I imagine buyer education becomes a huge, huge part of this we find across the board, across the 57 different business models that we track in the data center industry. Buyers are just simply not educated enough about a variety of topics, especially next generation power infrastructure. So what do you do? How do you educate buyers before the sales conversations even begin to set your team up for success? Well, I remember back at the beginning of this interview we said that the biggest problem is cutting through the noise, right? And what we have found is that education doesn't work if you can't get to the person who needs to be educated. And so what we said was we had to find out who the buyer really was. Right. And for us, the buyer isn't a technical buyer. So there's not really a whole lot of education to do in that space. Our buyer is actually at the real estate level. The person who's managing the real estate for the data center company, whether that's a dedicated data center company like an Equinix or whether that's a hyperscaler, they usually have a real estate team that is deciding on sites and figuring out that. And what they want to do is they want to know that when they sign on site they're going to hire somebody to build that plant, an epc. Right? Well, we take that role, we take on all that risk and then so we just go to them and we bond and insure our delivery capability to that. And then we take over and manage and operate the utilities beyond that. So we provide them with all the cooling, all the power, all the water management and all of that's inclusive into inside their lease. So they're getting a build to suit leaseback contract from us that includes energy. And so they don't need to pay the tariffs or the state. We don't need to wait for the public utility boards. We shave 18 to 36 months off of the hearing process because there simply aren't any. Right. We're not connecting to utilities, we're not connecting to the water, we're not connecting to the power. We're a zero liquid discharge facility and we don't emit anything. So guess what, a lot less permitting. And as a result, we can get these things done at gigawatt scale in like 18 months, which is a tremendous difference from the old way. And at the end of the day, you're not pumping a whole bunch of nox into the atmosphere. So with all of that in mind, given figuring out who the real stakeholders are, who actually influences who to spend time with this, what have you found works surprisingly well for energy infrastructure in terms of building consensus in terms of getting the right people on board, in terms of making sure that even the adjacent stakeholders learn what they need to know about this. Well, that's a great question because you know you're covering a lot of ground when you're talking to these buyers. Typically it's not one person making the decision, right? You've got a technical team, you've got a power team, you've got a real estate team, you've got a public relations team, and they all Got to be in alignment, right? So the key is to have all the right answers when you come to that initial contact. But typically you're not even allowed to approach that initial contact. You're talking to some kind of gatekeeper in front of them. And we have found that trying to go down the technical gatekeeper path is not successful because we're eliminating most of the traditional advisory work that those technical gatekeepers would do because we're providing it on a guaranteed power, guaranteed cooling basis. And that's not part of their control anymore. Right. It's now out of their control. So we found much more success by going in through the real estate agents, the trusted advisors that are coming in, speaking directly to the real estate teams. And then we educate the those people on the offering that we have and then they're like, oh my God, this shaves off so much technical know how that needs to be there. I can eliminate half of the sales process just, and just take this to the real estate team. And that has been very successful for us because once we get in front of the real estate person and they realize that all the pain and suffering that they have to do with all this technical advisory work because they're getting a very technical proposal goes away and all of a sudden it looks like a traditional, you know, take or pay real estate contract, which they're very familiar with. Right. They build those things all day long and they're looking for that guarantee to come from the, from the service provider like us that you're going to be on time and if you're not in time, you're going to pay us and there's going to be some kind of big insurance company involved from a bonding perspective. We had to go get that insurance, right? So we had to go convince the insurance companies that we were capable of doing this. And we were ultimately able to do so with the help of our engineering teams. They're also big guys. And convince them that not only do we know what we're doing, but, but that these solutions were worth bonding and insuring. And then on top of that, because this is sort of new territory for everyone at this gigawatt scale, we had to get technology risk insurance. Now this is probably our single biggest differentiator and probably the thing that is missing from every single proposal out there that's not coming from open origin is that we fully insure up to tens of billions of dollars the full cost of the full replace cost of the data center with our technology reinsurance program. And what that means is that if something Fails through our thing, let's say our superconductors fail because people consider that risky even though it's 30 years old. If that were not to work or we had to change the vendor, the insurance company will come in, pay for the entire replacements, cover all the liquidated damages to the customer while that's not functioning. So lost revenue, business interruption insurance, the whole nine yards, fix it, right? And then bring in a new vendor and repair all the stuff that needs to be done and get it working back on stack again all on the insurance company's nickel. Right? So we may not have $3 billion to come in and completely replace the micro grid on a 10 billion dollar data center, but the insurance company does and they're willing to stand behind it with that level of insurance because we went and worked with them to get that available. So we've created this new stack of extended technology risk insurance that simply is not available publicly on the market. What's fascinating about that is how you did that without the resources of a Fortune 500 company. Because I've had some people on the show that talked about getting into the managed services business and getting into the financial services side of this business, but they're coming from the perspective of having huge resources. How did you, as a founder CEO of a company with open origin, what, a couple dozen, couple hundred contractors all together, how did you make all this happen? Well, it depends on who your contractors are, right? So the key is, so, so my background is in very large management consulting, right? So all the companies that I've ever worked for are giant banks and insurance companies and, and those folks. So I had, I did have some context in there. But more importantly, when you're a management, management consultant at that level, so the Deloitte level of the world, right? And you see all the stuff that's happening, you learn about how these companies do this business. And when you see 50 of them, you realize they're all doing it pretty much the same way. Right? The difference between a startup and a Fortune 500, Fortune 500 company is nothing more than established process and procedure. That's it. So once you know what those processes and procedures are and you apply them to a startup, they work just as well at the startup level as they do there. And then you go to those really big, you know, white shoe law firms and engineering companies and things like that, and you speak their language and you start talking about technology, risk, insurance and they realize, oh my God, this, this guy knows what he's talking about. He knows how we do business. And then they're willing to step in. I went out and built a consortium of companies that we ultimately named the Energy Catalyst City Consortium, the eccc. It's hard to say, but it's, it's fun. And we have blue chip companies that are members of that organization that have come in and dedicated their time and put their own dollars in engineering effort into creating the shared intellectual property. Picture this for a minute. I got three or four big Fortune 500 companies to come in and share intellectual property on in the group that they then can work with on other sites and take back home with them because they helped create it. And the cross pollination created entirely new products for some of these companies. But we have our superconducting provider ended up creating a product that was uses 25% less cable as a result of some input from the solar team and the battery team and the engineering team that was helping us design the up. So an entirely new product set came out of that and then they're now selling that in the world. And that was developed about two years ago. But that's what came out of these companies being willing to work together. Now I didn't pay them to do that. They put their own time and engineering into effort because when I presented to them the idea, what we presented was so attractive that they saw that there was an immediate opportunity for them to be leading in the world in this area and it made sense for them to make that investment. What's interesting is across the IT industry for the last at least five years been talking about the idea of ecosystem orchestration. I'm sure all the market research companies have made a lot of big deal about explaining what that actually is. But that's essentially what you're doing. You're going out in the marketplace and looking at these gaps. And I imagine having access to generative AI tools helps people figure all of this stuff out faster. But it's still building relationships, it's still consensus building. And it's pretty amazing to watch all of that come together. Oh absolutely. And I will say one thing. So my very first version of the UPP took me 14 months to build the model because it is a bunch of linear algebra all mushed together. And because it's too complicated to model out in a Mathematica program, I had to build spreadsheets with approximations, right. And balancing the solar and the battery and hydrogen production and backup generation capabilities and all those wonderful things, right. To try and optimize the solution. Building that model the first time took 14 months. Now I Update the model by going to the AI programs and saying, add 12 megawatts of BESS and three more electrolyzers and 50 more megawatts of solar and come back to me with the delta and cost and see how it moves my level as cost of energy over the next five years. And in five minutes I have my answer right. And that, that has really phenomenally changed the game. We went from needing a team of 70 people working around the clock for 30 days to put together a $10 billion proposal because we bid on the Stargate project back when that was announced back in the day. And, and we, and we went from, you know, deciding to bid to finished production in 30 business days, which was phenomenal in and of itself for a $17 billion project with sign offs at the CEO level by all of our vendors. But we went from doing from that level of effort required to now I can produce a better estimate with more accuracy, plus or minus about 10% now instead of the 20% that we were looking at in that number, all with just the AI. Honestly, when you look at your particular sector of the data center marketplace, especially energy infrastructure, what go to market motions, do you see not working especially well in some cases, maybe even failing that, no one's talking about quite yet. Where do you see that? Like when you look at others in your space where you feel like, gosh, they must be really spinning their wheels. Yeah, I mean, there's a, there's a lot of wheel spinning going on. So that's like I'm trying to decide which of the things is really falling down. What's not working is building ahead of the demand before you've nailed down the client. So I think the most egregious example of this is former Governor Perry's extravaganza, let's call it Data center adventuring in West Texas. Right. I went out and they started construction. They had a buyer lined up. The buyer bailed in December of last year and they were not able to replace them, but they kept building. Right. So now they've got something half built, one third built, and they had to stop. And I think that's the biggest danger is you can get into a dedicated scenario where you've committed yourself to a project and you got to move forward because your investors are saying, hey, you got to do this. But on the other side of it, you know, the, the commitment wasn't quite there yet on the other end. And so he jumped the gun a bit, expecting that there'd be five other companies that could jump in. Right. Behind it that would take the space. And I think what we've all learned now is having seen several of those, those cancellations happen, most particularly, I think, when I believe it was OpenAI who jumped out of the Caruso project, and then Microsoft ultimately jumped in after Oracle, you know, took part of it as well. It can happen, but you got a gap, right? So if it's up and running and completed, then somebody will step in. But if it's half done, the chances of someone coming in and taking that particular one, that's very high risk. You, you're putting a lot of money on the table, in some cases, you know, hundreds of millions to a billion plus to, you know, break ground on one of these large facilities to get to the point where, you know, you're sort of trapped now on that site. And I would say that's the piece that's not working, is for those folks who kind of jumped the gun a little bit and thought they had one in the bag and didn't quite have it there. Yeah, it's really interesting to see the kind of advanced planning and making sure that it's truly a solid doing the diligence before moving ahead with these big investments. One other area I wanted to ask you about today as well as most data center GTM teams are flying completely blind. 83% of your buyer's journey is happening before they even speak with someone from your sales team. The Data Center Go to Market podcast is powered by DCSMI. We've studied over 1900 industry leaders to build a diagnostic framework that identifies exactly, exactly why deals fall apart and revenue stalls. Stop guessing and start benchmarking. Subscribe now to our weekly Data Center GTM briefing at www.dcsmi.com briefing. Again, that's dcsmi.com forward/briefing. B R I E F I N G. We know increasingly the technical buyers are doing a lot more research before you get to meet with them the first time. Ten years ago, they'd get like halfway there and now they're like 80 or 90% of the way there, especially because of all these great generative AI tools that they have with them. I mean, blogging, YouTube, LinkedIn, all the webinars. It's been around for 20 years at this point. Where do you think that technical content often fails to drive successful sales momentum, especially in and around data center energy? Oh, I think that that's a huge mistake thinking that technology is the thing driving the conversation. It's not. I think that's the biggest mistake. Probably anybody selling data Centers today probably could be running into from their sales department. The technical magic for your facility doesn't matter at all. There's so much demand right now. The only thing that matters is speed to power. How fast can you get it done? It doesn't matter if you're more efficient or if the, if you got a PUE that's 10% better than your competitors or that you're even, that you're cheaper, that doesn't matter as much. What matters is can you get it done? Because the amount of money that these AI companies make by being live one month sooner, right. In some cases can be 10% to even 50% of the investment capital cost of the project. That is game changing, right. Only thing that matters is speed. So the, if you can commit to speed and you can get in front of the buyer who, like as I mentioned before, is probably the real estate team that ultimately makes the final decision and you can streamline that process and show that you're faster on the permitting, you're faster on the, on the layout, you're faster on the deployment and you can get something up and running to 2 or 300 megawatts, which is meaningful to them in like a 8 to 12 month window. If you can do those things right and then ultimately scale to a gigawatt within 18 months or less, that's the ante to play. If you can't do that, you're spinning your wheels and no one cares about your pua. They just don't. No one cares about your pollution. I wish they did. I would sell. I would beat every competitor if that was the case. Maybe in Europe, but yeah, yeah, in the United States, in Texas, no one cares. It's like how fast can you get that Watt into my machine, right? And you got to ante up, right. And, and, and deliver that. And if you're not willing to guarantee it, then you're going to get passed by, right? Because oh, they might be interested like, oh great, you could do it in 12 months or you can do it in 6 months. Fantastic. Are you willing to bond to that? Nope. Well, sorry, thank you for playing. Right. And go on down the road. So, so that's really the key is not only do you need to be able to be actually able to do that fast enough, which is really hard to do unless you're the biggest players, but you also have to be able to be willing to prove that, put money on the table. And if you are unwilling or unable to put a billion dollars down on your $10 billion engagement to bet the farm that you can deliver on yourself, then the big guys aren't going to give you the time of day because that's what they require now of the, of the big EPCs. So in terms of the technical credibility that gives the buyer confidence, it's not just speed, it's a lot of financial services, it's a lot of risk management, it's a lot of basic insurance issues. So this is a CFO decision, chief investment officer decision. Who's typically the ultimate, who's, who's running the committee? That's a great question. Right, so this is, so this really comes down to how you're selling the solution too. If you're expecting a hyperscaler to walk up and buy a multi billion dollar data center, you probably should be in a different business because that's not how they do it. Right? At this point in time they are not looking to put a $10 billion boat anchor on their balance sheet. Right. The hyperscaler wants to lease and the reason they want to lease is because they know in five years they're going to abandon that lease, most likely because that data center will no longer be in alignment with reality and they're, they'll backfill for a while, they might sign a tenure, but really you're probably pressing it at that point in time. There are a few people out there where they're building things themselves, right? We know that Meta is in that space where they've taken some things on directly, but that was their last resort, right? They didn't want to do that. It's just the EPCs all said, no, we're not taking that risk. So they had to do it themselves. We removed that from the game by making it a, a lease based program. Right now there are big companies out there that are willing to do this, right? All the big real estate guys are willing to fund these big things as long as the off taker is one of the hyperscalers. So getting the funding for the project isn't the problem. Convincing the funding guy that you're the guy to bring that to the table, that's really the key. And that was the other thing that we had to do at open origin in order to be able to play in the space is we had to go convince one of the big guys that we were the right answer to come in and play in their portfolio and be part of their solution set to the hyperscalers. And so we did. We got a big funder behind us. And so now what we offer to the customer is an all in lease, right? And I think that that really is an innovative offering because today you can get the lease from these guys exactly the same way. But it doesn't include the power, it doesn't include the energy cost, it doesn't include the cooling cost. All those are extra and need to be somehow managed by the EPC. So 20% of your power, 15% of your power is going to go towards cooling, right? Well, we looked at that problem and said the whole thing should be part of one solution. Right? Integrated the whole thing. So you know, there's a differentiating factor in what we're delivering. But more importantly, what you're delivering has to be like, it has to look and feel like an operating cost to the hyperscaler because that's, the hyperscaler is burning cash on this, on the GPUs, right? They have to make this huge investment and people are like, oh my God, they're spending billions on the data centers. Guys, the data center is not the big cost. The big cost is the GPUs that go into the data center that are only good for 36 months, right? They're, they're, you know, into my $10 billion CapEx data center, you know, at sort of gigawatt plus scale, they're now putting $50 billion worth of GPUs. All of a sudden the real estate is not the expensive piece, it's the GPUs. How have you had to construct your own commercial team, your own go to market team? Increasingly across the industry we're seeing more and more of the technology companies hiring career IT professionals who never worked in a commercial role, sales role before. And companies like yours are increasingly going to engineers. What's the profile? Who's the right person that has the credibility to be a peer in these kinds of conversations? We take folks right out of the consulting world because if you think about it, folks like myself who came out of management consulting at that, so the top tier firms, we sell the future, right? That's what a consultant actually does. They're not selling a product, they're not selling a service. When you walk in and you're a management consultant, you're talking directly to the CEO of a Fortune 500 company and describing to them how they're going to save money or make more money or whatever it is that you're tackling for them, right? You literally have to understand fully what they're doing, understand fully what the market's doing and then you creating a solution out of whole cloth for them. That's brand new. That's going to make a difference based upon all the best practices that you've seen across the industry. That's exactly what's happening right now. And that's exactly what we did at open origin. That's not being done by most of these data center providers. They're taking business as usual what we did on the grid and we're trying to turn it into something that sort of fits this new, you know, hyperscaler world where everything is so much larger. But guess what? Cooling systems don't scale the gigawatt scale. The same thing that works at 100 megawatts does not work very well and creates a huge heat island effect out out there in the world. If you do it at gigawatt scale and if you take it to 9 or 10 gigawatts like we're doing out in Utah, when Kevin O' Leary's scenario, you're creating massive heat islands with the amount of heat coming off of these things, that is unprecedented in human history. We've never done anything that hot all the time. So we really don't know what the effect is. And so these are the types of considerations that simply aren't present anywhere else. And it doesn't work to take to assume that doing something that worked at 100 megawatts and scaling it up to gigawatt plus is going to be the right answer. So I think that's the thing. I'm seeing that where there's gaps and it's probably not the right sales guy to go out and hire the guy from the utilities industry to do this because he's thinking about it completely differently than the management consultant would. So it's very different than a traditional account executive, account manager, very different playbook band kind of mindset with all of this. And what's interesting too is you mentioned earlier that there are certain kinds of stakeholders that were basically seeing their jobs threatened because you're doing something that's reasonably disrupted. So that challenges the whole idea of talent shortages that so many people are running around complaining about and terrified about in the data center industry. And yet there's a completely different kind of talent that you need to be able to advance something that's this different than the perspective that real estate developers, real estate investment trusts are used to thinking about. That is very, very true. And to kind of put a really sharp pin on that, we thought we had a great partnership arranged with one of the top technology advisors around energy for the data center industry. Right. They were folks that had advised one of the largest hyperscalers in the world for two decades had helped them create what an AI data center looked like. And their partner was very excited about our solution. We had had several great conversations and so then they brought in their technical team because their technical team were owners in the partnership to vet our solution. And the technical team laughed us out of the room. Despite the fact that our solution had been designed by the number one data center engineering firm in the world for the past five years, it wasn't open Origin saying this is the thing that worked. This was the number one engineering firm in the world saying that this solution works. And it had presented at Cirid and it had won accolades and been presented twice. That's how good it was to hundreds of engineers at SIR IT who had held it up as the example for this is how we should build engineering the future. And these guys were so set in their way that they could not see the forest for the trees. And they left us out of the room and the partner was so embarrassed that the partnership fell apart. It was really crazy. And now. But effectively what, what those engineers saw was they were completely irrelevant to the conversation. They were just, they had no job. Their job was just gone. It wasn't, it wasn't, you know, fading into the background, but their entire reason for existing as a consulting firm was disappeared. Reminds me of Moneyball threatening their whole way of life. The old fashioned scouts that didn't want to use analytics in this and spreadsheets. That's exactly what it was. And, and they, they weren't using AI that way. They were still doing the same thing. They were still convinced that, you know, you know, things had to be built in a certain way and the hot and cold racks and the whole nine yards and you know, even though they, they'd done things that were very innovative earlier, right. They still were thinking about it from a technology perspective. When it became a real estate transaction, they didn't know how to handle that. It didn't make sense to them anymore. Now the crazy part was, is that we still would have hired them. Right. They had value to us as the customer. But what they saw is that they were moving out of the control position and into becoming a technology advisory service to the actual service provider. And they couldn't stomach that because they were, they're making top dollar right now, right. Thousand dollars an hour rates. And now we're going to be, now they're going to just be also ran engineers, you know, at normal engineering rates, advising a real estate company on building stuff. And they, they Just couldn't find themselves in that position. Yeah, we're going through the exact same transition. When we look at commercial teams, we're in a world where IT buyers and engineers are getting their answers directly from their favorite chatgpt Claude, what have you. The account executive and the sales development rep and account manager would be phenomenal. Curators, moderators, explainers, bringing together all of these subject matter experts. But they're going to have to sit in the co pil seat next to the true subject matter expert to get back in the door with the peers because nobody needs them to read the information on website or YouTube or on the generative AI tools anymore. So yeah, for sure. And it's got massive change, massive implications too, even in this industry because hundreds of millions are spent annually on these massive conferences and exactly the buying habits have changed so much. Well, and on top of that, and I'll just tell you internally, right, you know we, we have a really tight chip on the, on the go to market side. We actually have three dedicated resources only for all of the work that we do. And that's not just in the data center space. We actually serve the chemical and manufacturing industry as well. So we have a, we have a partner manager who manages all of our real estate partnerships. And those folks, as we said, that's our, you know, really the path to get to the, to the decision makers through the real estate side. That's where we found is very successful. The, we have a real, actual dedicated salesperson for federal, right? So all government, we have a dedicated person because that person, you got to have the dedicated federal person. That's how that's going to work. And they've done really well because they've been super dedicated just to federal and then everything else goes directly to the CRO and the CRO personally manages because he came from the big guys, he's been CRO for, you know, Tata and those guys, you know, really big consulting firms. And the reason for that is the volume of, of actual qualified leads isn't very big. It's a few hundred qualified leads a year. And the decision time on a particular lead is measured in months. Right? So having, you know, if you can get three or four phone calls in a month on a particular account to move things forward as things are getting hot and heavy, you're doing pretty well because the hyperscalers love to make you hurry up and wait. Oh my God, they love that so much. Right? Because their procedure for going through this historically is it took five years from hey, we need to build Another data center to, oh, let's break ground, right? They want that now to be shortened, but they're not playing like they want that to be shortened. Right? They're not acting like they want this to take 18 months. They're still taking 18 months to get to an 18 month decision instead of taking 2 months to get to an 18 month Decision, which is actually where we could be if they simply said, hey, look, companies like Open Origin, you're covering the financial risk, you're providing the land, you're providing the facility, you're providing the utilities and everything, which is what many of these, you know, companies are now offering. I'll withhold judgment on whether they actually can deliver all that, but that's beside the point. And if that's really the case, why am I doing all this other stuff? Why am I taking 18 months to get to a decision? I mean, don't I want my GPUs to be deployed? I mean, didn't, didn't Microsoft's CEO very recently come out and say, I've got warehouses full of GPUs I can't deploy because my, my data centers aren't ready yet, and yet it still takes 18 months to get to a buying decision. I think there's definitely stakeholders in different pockets of this industry that feel like if they can keep the status quo going, no one will notice that they may not be as valuable as they were 12, 24, 36 months ago. And they won't have to reskill, move around to a different part of the industry, confront the reality of how quickly the world is changing. I'm going to say unqualified, yes, that's absolutely the case. However, it's unlike most of the conspiracy theorists, it is not a coordinated effort. Right. It's individually they're doing that, but collectively the effect is very similar. Right. You still got people who are playing like, who are on the technical side, who are like poo pooing the solutions that are being offered and being like, no, there's no way that can be true because of X, Y and Z not realizing that X, Y and Z aren't even applicable in this case because they're not grid connected or they're using a different battery technology or they're using superconductors so there's no losses in the system. So all these assumptions that they've made their living on working around no longer are even there. Right. Like 24 months ago it was, oh, nobody can build power in a data center in six months until XAI did it Right. And prove that you could just by schlepping in a bunch of batteries and generators into a room and calling it a day. Right. And running everything on low power because you couldn't get a transformer in the door. Now, is that the right way to do it? Probably not. But is it a way to do it? Yeah, it did. Now, is he paying through the nose? I mean, he's probably paying the equivalent of, you know, 20 or 30 cents a kilowatt hour for all of that because of the. But he's. But all that's internal cost. Right? I mean, the guy owns a battery factory, so, you know, he was able to do things like bring in, you know, a gigawatt hour plus worth of batteries to support and even out his flow so that he doesn't knock. So it doesn't knock his grid generation offline every time there's a spin up on the GPUs. So I think there's new challenges here, you know, on the data center side with large jobs going up and down. You know, no one's ever seen a gigawatt of GPUs turn on and off with an off switch before. Right. In reality. And yet that's exactly what's about to happen. And we've already seen that the grid's not real friendly for that kind of load. Load swap. Right. 900 megawatts on and off. They don't like that in half a second. Right. That is not a normal load profile from the grid's perspective. And it flips breakers sometimes. Right. But you can't predict when it's going to flip and when it's not because of the strain on the system. So something that's perfectly safe in January, in June, not so much. That's what happened in Spain. We had, you know, we had a failing back there last summer that caused that big, you know, massive outage that cascaded all across Europe. And that's not. And we came within one breaker failing in Virginia two years ago that a lot of people don't know about. But there was a similar incident that almost occurred where we had one breaker that didn't blip. And if it had blipped, the whole east coast would have been out again, just like in Spain. So we came within a. Within the. The quantum tolerance of a transformer, not a circuit breaker not flipping on. Just because we lucked out. Right. I don't know if I want to bet the farm of the entire east coast again on whether or not a data center breaker holds. We're going to do a follow up interview with I had someone on the podcast a month or two ago who is a 35, 40 year retired veteran Dominion Energy guy who spent his entire career building out electric infrastructure for data centers in Virginia. Be really interesting to get his take. Yeah, I'm sure he'd be like we did not plan for this is probably what he would be saying because that to put it in perspective, two years ago all of the Virginia D.C. area, you know, Tri State area data center capacity combined was 2.3 gigawatts two years ago and today we've got people building individual data centers that are twice that size. That's crazy. Insane growth. Yeah, insane growth. Final area I want to ask you about today Wesley, is when you look into your crystal ball, thinking ahead 12, 24 months from now, what do you envision changing for a company like Open Origin, what do you think is going to surprise the industry the most? I think the market is going to come around to a lot more renewable instead of natural gas and I think that the reason for that is going to be surprising. I believe that the reason for that is that natural gas is going to cost five to eight times as much as it does today in West Texas because of the war and the crisis going on over there. The LNG exports are going to increase the cost of natural gas. That's going to increase the cost to the customer in in Texas who's currently buying unprecedented low prices in the natural gas space and when all of a sudden that cost 10x what it did the day before yesterday off grid natural gas power is not going to look nearly as attractive as it did previously and people are going to take a really hard look at solar in particular and wind possibly combined. Although wind takes a lot more time to set up, solar is very fast. So I think we're going to see a whole lot of market alignment around solar with backup thermal generation as where solar is the primary and batteries because the battery prices are continuing to drop. There's new tech hitting the market right now with sodium batteries for example, that the full expectation is by the time a sodium battery hits scale for consumption in the United States in China it's going to be 20, $20,000amegawatt hour, which is about a third of the cheapest LFP battery is today. And at that price solar and battery is unbeatable by any fossil fuel ever. In fact, if the fossil is free it's still cheaper because the the cost of operating the turbine and the capex recovery on the turbine is bigger than the cost of the solar. So the company that I spoke with last year that can't get turbines to be manufactured fast enough to keep up with demand, it will likely at some point level off because the cost of the gas is. They'll be able to get the gas, but the cost of the gas will be too expensive, no matter how good they figure out the locations and how matter much they catch up with the manufacturing of the turbines. Well, remember what I said earlier, the only thing that matters, the speed to power. Yeah. And it doesn't matter what the cost is right now. Right. Because the upside is measured in the billions per month. Right. Of upside to cost. So what matters is the increasing automation of the solar outlay. So we're working very closely with robotics and automation to maximize very dense solar outlays. When we started, even with robots and really detailed cruise you could do, took about a week, about two weeks to do a megawatt with four guys and maybe a robot helping them. Right. To help unload stuff real quick. That has accelerated now and we're currently anticipating that in the new models of robots that are just being released right now and we're going to be start testing this summer that we should be able to do with three robots and four guys, 10 megawatts of deployment in a 24 hour period. What's wild is how quickly this narrative has changed around solar. As recently as two years ago, people would come on and they'd say, yep, solar is a good idea. It's really good for pr, it looks good for sustainability. It'll run the lights in the parking lot and the interior lights and that'll be about it. That the yields and the prices will come down massively enough. That this can fill up the, in the daytime, fill up the batteries is pretty remarkable. Yeah. And it's not like it doesn't take a lot of space because it does. I mean, let's not, let's not be wrong. But when you compare it to the combined footprint of the oil and gas industry for the harvesting of the oil and the gas fields and, and the refineries and all that great stuff together, it's comparable or better even with traditional sort of 3 acre or 4 acre per megawatt solar arrays. We start looking at dense solar though on fixed tilt axis. You know, I'll give you a little bit of insight. But right now in China you can deploy a megawatt of solar all in at about 18 to 20 cents a watt, which is absolutely phenomenal. In the United States we probably do it cheaper than anybody else out there. And we're currently in the sort of 26, 27, 28 cents, a lot depending on the brands. And that's because of the tariffs. If we did it in Mexico, we could actually do it, believe it or not, cheaper than to do it in China because we of the racking methodologies that we're using because it's been so expensive. One of the things that we have an advantage over the Chinese folks is that the racking technologies and the speed and the robotics have been forced to automate faster in the United States on the deployment side because they had plenty of labor that was super cheap on the other side. So that's an area where they've innovated in technology that allows you to pick up a panel with a drone or something and carry it to where it needs to be and give it to the guy who then installs it. Right. America had to automate all the way to the rack from the pallet on the truck. Right. And so we've got robots now that pick up the pallets and take them out to the field. And there's another robot who unpacks it, there's another robot who puts it in place. And then now we've just got a guy who just goes zoop, zoop, zoop, zoop, puts in four screws and moves on to the next, the next one. And at. With 800 watt panels, which are now the sort of the standard HJT panels coming out of China as the latest, greatest version, soon we'll see a thousand watt panels, no problem at all. By the, by the end of next year, we anticipate 1,000 watt panels to be the de facto standard. You're decreasing the wiring costs, you're decreasing the speed to install it, to install all this stuff. And at that point, we should see US prices in the sort of 22, 23 cents. A lot. Assuming that the tariffs stay as terrible as they are today. And in Mexico, where we're also looking at some stuff, we should be able to do it in Mexico for 15 to 18 cents a watt. So if this picks up really well in industrial and hyperscale environments, will this go pretty mainstream too, in regular commercial residential adoption? I've lived in the Sunshine State for 25 years, and I, when I come up and visit New Jersey, I see more solar panels around residential neighborhoods than I do in Florida. Yeah, I mean, I hope that it does, but the, the reality is, though, is that the innovation that's happening on the utility side of the house is so Far in advance of the residential side. I mean, and it just goes to show you that if you go to the big guys and you look at the type of panel they're selling to the utility providers versus the type of panel they're selling to the homeowners, it's night and day. The panel of the homeowner is three year old tech. What they're trying to do is they're trying to get their money back from investments they made five years ago into technology that nobody's buying anymore in that space. Right. They move so fast that they thought they were going to get five years worth of production. They got 18 months and then the next version came out. And then they got 18 months and the next version came out. Then they figured out now that they only get 18 to 24 months per run. And so they're not making the huge factory bets that they were before on the chip, on the, on the chip side of the sell side manufacturing. So that has helped, but only partially. So now the runs that are going now, they're now modular trains and they're building like two gigawatt trains. And the reason I can say this with confidence is some of our customers, remember we serve that manufacturing industry as well. So we're serving some of the very large solarvoltaic customers and looking at their US manufacturing facilities and helping them design and build those things out. And so they're now thinking about them in two gigawatt trains and inside a much larger, let's say 10 gigawatt footprint. And then they rotate. So we would provide all them, their power and their infrastructure just like they were a data center because they need that kind of reliable power for their cell production. And then, then they automate the assembly. It's a lights out factory. There are zero employees inside. During the day it's dark and it runs. And then basically the logistics guys in the take the, take the final packaged solar panel and put it on the truck. Like that's where the automation sort of ends right now in the United States. But we expect, and we're starting to see now some automation on the warehousing side that hasn't been there in the past. And also on the construction side, sort of the two areas that we thought would be sort of the last to be automated, we're now starting to see those being automated. So we're very excited about potentially looking at the 10 person factory right now. Right now we're looking at giant factories that have like 100, 150 people. And you know, we, but we fully anticipate that that's going to cut down by another factor of 10 within the next two to three years so that you'd have this like sort of 10 gigawatt solar facility with like 25, 30 employees. It's another reason, I imagine leasing is going to be a big deal, because it's not just that the life cycle of the GPUs moves much faster than the real estate. The life cycle of the solar farm moves much faster than the real estate. Well, the great part on the farm side is that once it's in place, you're sunk cost, right? So now you just keep on going. And it's so cheap now. I mean, the panels, I don't know if you guys have looked at the pricing and panels. If you're buying them directly in China FOB before the tariffs, you're looking at 8, 9 cents a watt for latest, greatest off the shelf panels. And these are not discounted because they're cheap. These are being sold by the tier two providers who have spare capacity because the tier one providers, you know, are maxed out and so they've got extra capacity in their lines and they'll sell you that, that extra capacity at a price point that's absolutely ludicrous. Your levelized cost of electricity when you're in your solar panel is that cheap. And your fully installed cost is like 28 cents. Turns out to be about $9amegawatt hour. That's less than a penny per kilowatt hour, right? When energy is that cheap, who cares what, how much, how efficient you are at the data center level? It just doesn't matter. Your PUE could be 1.4. Who doesn't matter? It costs you. You know, we're literally giving away the power. We don't charge for it because it's so cheap. We just charge for the capacity. This has really been an eye opener, enlightening. Talking about a lot of areas cutting across energy infrastructure, real estate, decision process disruption. How to explain this rethinking the business models. Really, really enjoyed having you on the podcast, Wesley. If someone wants to follow what you're working on or reach out with you or learn more about open origin, is LinkedIn a good place to start? Yeah, LinkedIn's great. You can reach me on there. WesleyPowell and then also OpenOrigin Industries is our website address and that's probably the best place to come see our latest creative stuff. You can see all of our support lines right now, which includes AI fuels and also manufacturing. And we're in the process of actually updating that to expand that offering so people can get a better idea of how we provide those services. And imagine by the conversations we had today that education is a huge, huge part of what you do. It really is. But funnily enough, it happens after the sale now instead of or after the initial sales conversation. People come to us because they're interested in cheap power for their data center, and then they're like, wow, we can buy the whole thing from you. And I'm like, yeah, that's the whole point. That's so cool. Yeah. And speaking today with Wesley Powell, who is the CEO and founder of OpenOrigin. Thanks for joining, Wesley. Thanks so much, Joshua. Most data center GTM teams are flying completely blind. 83% of your buyer's journey is happening before they even speak with someone from your sales team. The data Center Go to Market podcast is powered by DCSMI. We've studied over 1900 industry leaders to build a diagnostic framework that identifies exactly why deals fall apart and revenue stalls. Stop guessing and start benchmarking. Subscribe now to our Weekly Data Center GTM Briefing at www.dcsmi.com briefing. Again, that's D C-S-M I.com briefing B R I E F I N G.