[[{“value”:”
This is the first in a new series for 2025, in which we will discuss how AI impacts energy throughout the entire supply chain and on to consumers.
In this episode of Energy Newsbeat – Conversations in Energy, Stuart Turley hosts experts Mark Stansberry and Jon Brewton to discuss the intersection of AI, energy, and business. Mark, a seasoned oil and gas professional, shares his excitement about the potential of AI in the industry. At the same time, Jon, the founder of Data Squared, explains how his company is tackling the challenge of connecting disparate data systems. They explore the importance of explainable AI and its application in high-stakes sectors like defense and energy. The discussion also highlights the role of AI in optimizing costs and improving decision-making across industries.
Thank you, Mark and Jon, for stopping by the podcast! This was an important episode, and it will be an excellent start for the series.
Please follow Jon on his LinkedIn here: https://www.linkedin.com/in/jon-brewton-datasquared/
And Data2 website
https://data2.ai/
Check out everything about Mark Stansberry here: https://www.linkedin.com/in/mark-a-stansberry-67407511/
And his website:https://www.markstansberry.com/
Highlights of the Podcast
00:00 – Intro
00:44 – Guest Introduction: Mark Stansberry
01:03 – Guest Introduction: John Brewton
01:21 – Mark’s Role as Advisor
02:40 – John’s Company, Data Squared
06:00 – Solving Industry Problems with AI
08:43 – Business Application: Connecting Disparate Data Systems
14:07 – Example of Solving Data Challenges
18:32 – Value Chain Optimization
21:08 – AI in Engineering and Strategic Planning
25:03 – Overcoming the Challenge of Hallucinations in AI
32:00 – The Role of AI in Federal Government and Energy
33:03 – Collaboration with Louisiana State University
35:50 – The Future of AI and Energy
39:01 – Closing Remarks
Stuart Turley [00:00:07] Hello everybody, welcome to the Energy Newsbeat Podcast, my name is Stu Turley, President and CEO of the Sandstone Group. We are living in an absolutely wild time. And I mean, AI is changing and if you don’t understand what you’re looking at, how do you know if it’s going to do you any good? I need experts and I have two experts today that I rely on. And I’m telling you right now, Mark Stansberry is a friend of the show. He has been on the podcast and he has written a book, American needs energy and it’s natural resources. Thank you, Mark, for stopping by today.
Mark Stansberry [00:00:46] Thank you Stu Appreciate this.
Stuart Turley [00:00:48] And we’ve gotten John Bruton. John Brutan is the founder and CEO. He is a United States Air Force vet, MBA, went to Harvard Business School, and he is the Founder and CEO of Data Squared. Welcome, John. How are you today?
Jon Brewton [00:01:03] Very good, sir. Thank you for the time. Look forward to the conversation.
Stuart Turley [00:01:06] I’ll tell you, we’ve had a little warm-up discussion here, and I’m really looking forward to hearing about this, but Mark, you were promoted, I guess, or made an advisor to the CEO, John Bruton. Tell us about that.
Mark Stansberry [00:01:22] Well, I had a mutual friend that introduced me to Data Squared, and it was in the technology world as well. And so having been in the oil and gas business for around 48 years, AI, like you’re talking about, Stu, is a new venture and really exciting for someone like myself. At this stage in my life, I’m excited about the future of energy, and AI is going to be a big part of that. So I had the opportunity to team up. With Data Squared and John Bruton, who’s the CEO, strategic advisor to the CEO John Brunton, that’s J-O-N-E-R-EW-T-O N. And his company has four partners that are from the military background, but also have unbelievable experience in the energy front. And I will let them, or let John, if he will, tell more about where it began and his journey as well. I think that’s important, Stu. To hear about that because it’s one thing to have the background that I have and of course that you have as well, but to now be placed in front of this AI with I get calls all the time now about data centers, about how we can incorporate AI, and I can’t think of any other company or person that could lead this discussion than John. So John, if you will, floor is yours if that’s All right, Stu.
Jon Brewton [00:02:40] Absolutely, sure. Thanks for the kind introduction there, Mark. Yeah, I’m John Bruden, founder and CEO of Data Squared. I’ll give you a quick intro to the company and kind of how we got to where we are and what we’re up to at the moment. We founded Data Squired in September of 2023 with a vision to build a world where people and machines could make smarter decisions together. Our team, as alluded to previously, includes co-founders Chris Rohrbach, who’s a former Navy SEAL commander, retired after 26 years. I saved him from a life of being probably a Fox News commentator or something else, given his background. He’s a fairly decorated war hero. We ended up spending a lot of time together, doing our MBA together. And so once I decided to start a company, I called Chris and said, Hey, what are you doing? And he said, well, I’m looking for a job. I’m about to retire. I said, I got one for you. So we, we started a company together with another friend of ours, Eric Costantini, who’s a Marine Corps veteran. And Eric and I actually studied together at Harvard Business School. And we also have our fourth founder who’s Jeff Doglish. Jeff is a oil and gas industry veteran, 30 years. Jeff started training logistics dispatchers on rigs on how to schedule things in the Gulf of Thailand. That was his first experience working for Unical and ended up working, running the entire upstream data and really sort of IT infrastructure for Chevron for a long time. So Jeff and I worked together at Chevron for several years and Jeff left Chevron a little bit before I did, started a different company. It was based on knowledge graphs. They were trying to attack how to build knowledge systems, aggregated intelligence systems for oil and gas. And they could never really figure out to operationalize it because it’s pretty hard to build and it’s pretty hard operationalize. And I’ll talk about why that’s important in just a moment. My personal background, US Air Force veteran spent time with both BP and Chevron 18 years in the industry for everywhere in the world, I think, in Antarctica, and kind of lived everywhere as well. Worked everything from an upstream to midstream perspective. Did basically everything. Production engineering, drilling engineering, facilities engineering, strategy, acquisitions to vestitures, production ops, like everything during my career. So my career in oil and gas is a little bit different. It wasn’t limited to one function or one type of role. I ended up doing a lot of different things. And there was a common problem that I faced regardless of what I was doing, where I was working, what I was working on very specifically was the sort of disconnected nature of intelligence within our industry. You know, everything is very functionally siloed at times, very hard to sort of connect in many instances and depending on which company you’re working for, you might have good relationships with your functional partners and you might not depending on how the organizations are set up. And so we kind of took our experiences from the oil and gas industry and our experiences from the intelligence space in DOD and the US government and try to figure out how we could take a different approach to solving problems. And it started with How can we aggregate intelligence in a really meaningful way so that we can connect things across these silos in a, really easy and really quick fashion. That was kind of the first idea that we had, but we started having discussions about how to build this company out right as we got to the release of the first chat GPT model. And again, timeline wise, I said, Hey, we started the company in September, 2023 will be, I think, hit. Commercial availability where people could get subscriptions in March of 23. And so we were trying to figure out how to aggregate intelligence in a really smart way, how we can make that easy for people and then how we could use some of these new capabilities that were on the market to build better tools and discover a better way to bring information together and to do analytics and to create a collaborative environment over the intelligence that we gather. And turns out that Jeff’s experience trying to build a knowledge Graph company, my experience trying to commercialize that company inside of Chevron and our experience with these tools, just in terms of AI and machine learning led us to picking a very good data model that interfaced with large language models, you know, Chad, GPT, Claude, Jim, and I, in a way that solves for really sort of three. Problems. One, you have problems associated to hallucinated content, you know, the way that these systems work, they tend to deliver an answer regardless of whether or not they have the right materials to actually provide that answer, because yeah, they work on a decision imperative, which is an objective function to produce an answer. That creates a little bit of a lack of trust. These systems themselves are not inherently multimodal. And what I mean by that is we can’t take spreadsheets and photos and written logs and any other sort of intelligence that we have PDF documents and put them in the same place and pass those to a large language model without it breaking. It just doesn’t work whenever you pass that sort of information over. And so we really tried to figure out how we could take the rigorous demands of what are these high stakes industries, both defense, intelligence, and then the oil and gas industry, and design something that would make it easy, and make it fast, and makes it hallucination-free, fully transparent, fully trustworthy, and explainable. Because that’s really what’s going to be required if we’re going to use it as an engineering assistant. Or if we’re going to use it in an intelligence capacity. That’s kind of how we started. And we, we used our experience from these really interesting domains that don’t sound like they would be connected very, very much. You know, the intelligence organization, say like the CIA actually in practical application looks a whole hell of a lot like Chevron internally. And so taking sort of these best practices from these different places and coming up with a new approach, that that’s how we got started. And I think we’re doing some pretty interesting things.
Stuart Turley [00:08:43] You know, this is very critical because the disparate business systems that are in the oil and gas space are archaic and very painful. And in being, when you’re doing looking at deal evaluations, or we’re looking at oil and gas flows, we’re saying, Hey, what are the offsetting wells and I’m trying to evaluate? Is this a good investment? Or is this a bad investment? I have to have data. It costs me a lot of money. If I don’t, let’s take it just as an sample grog. Grok, however you want to phrase it. I like it as a tool, but boy, you got to be careful because you mentioned that earlier that is systems have dreams or however, how did you phrase it? I love the way you
Jon Brewton [00:09:27] hallucinations, yeah.
Stuart Turley [00:09:27] Hallucinations thank you kind of like my ex-wife I mean it was absolutely horrific some of the questions some of it is absolutely phenomenal but you gotta understand enough of the material to go wait a minute I see something wrong let me drill a little bit into this how have you been able to stop the Groakism?
Jon Brewton [00:09:49] If you. Yeah. Yeah, that’s a perfectly fine way to say it for me. Like in general, like all these systems suffer from similar issues. And there was a great Wall Street Journal article. I think it was Wall Street Journal a couple of days ago, really got into the scale of these issues and some of the problems that folks were seeing. Sorry, it was a New York Times article and. The headline of this is AI is getting more powerful, but hallucinations are getting worse. A new wave of reasoning systems for companies like OpenAI is producing incorrect information more often. Even companies don’t know why. And I don’t want to get into why it happens, because there’s a lot of technicality behind how these systems are built, how they work and how they do what they do, which is irrelevant. All you need to know is that this problem persists, whether it’s grok or open AI or clog or Gemini, or if you built a model yourself internally, it wouldn’t matter because of the way that they’re built to actually produce the answers that they do. So we figured that out really early on. We figured it out in quite honestly about April of 2023. Even before we started the company, we figured out that. One, there’s some inherent problems associated to what is sort of happening with these systems. And if we prepared data in the right way, we could not only use multiple modes of data, you know, PDFs and Excel spreadsheets and visual representations of things simultaneously, but we could do it in a way where we could ground those models in the reality that we provided. So a lot of the problems that they face are from Drift, You know, because they work on human language. So whenever we ask a question. The value of the answer that is provided is really determined by the words that we use because it looks into the words that we used and it tries to connect it to the information it was trained on. Now some of these things are very domain specific, as you know, there’s probably a hundred words in the English language that are repurposed for other things in the oil field, you know? We could probably go down to laundromats right now or Texas, where I’m a hundred percent Louisiana has probably got us beat, but you know there’s this, this really weird thing that happens where the system tried to derive context from the words that we use. And so a lot of people think if we just. Use more words when we prompt the systems or more detailed in the way that we say stop that will help and it’s not true it kind of works on a bell curve the less word you use the higher likelihood of failure the more word you used higher likelihood failure. Somewhere in the middle is a bit of a sweet spot. Doesn’t make a lot of sense, but this is a scalability problem. If we, as an industry, especially, you know, this is a high reliability industry, if we make mistakes here and we’re using these tools, that can cost people their lives. And to your earlier point, it can cost us a lot of money, so we need to know, we need to know that these things are working correctly. So that’s how we kind of started. That’s the imperative we started trying to build around. And we figured out that if we prepare data in the right way, and we provide it to these systems in a different manner than a traditional sort of question in natural language, that we can essentially fix the way that they see the universe to that data. And so we can tell these systems, okay, here’s the world as you need to understand it. This is the world that you need understand. You don’t need to think about anything else. You don’ t need to try to make up anything beyond what we provided you. You can use the stuff that you were trained on to inform how you answer this question, but you can’t use anything from your training interjected into the answer that you provide. So that’s a bit of the kind of workflow for us. And then the analogy is when we prepare data, and we submit it to a large language model, we’re essentially putting a puzzle together. We take a thousand piece puzzle, and we put it together, and we present it in that format to these models and say, here is the world, here’s how it’s connected, here’s why it’s connect. And this really helps us to define the who’s, the what’s, the when’s, and the where’s. And so when we start asking questions, we can quickly get to and ascertain the how’s and the why’s associated to anything.
Stuart Turley [00:14:08] This is cool. Let me ask a real world business scenario. Let’s take 1500 pads over at BHP and these 1500 pads have a great SCADA system. They’ve absolutely been connected. We know their well log history and we know everything else and then they’re being sold from BHP to pioneer somebody else. Well, the one over here is running SolarWinds and they’re running their databases and everything else. So am I asking a question for some of your potential future customers? Would they be able to take data squared and say, I’ve got disparate business systems and I want to connect this skated data over here to this program over here and take a look at a completely different W operating accounting software and look at it over here.
Jon Brewton [00:15:00] I swear to God, we didn’t prepare this question in advance. You just, he teed me up in a great way. So it’s cause I’ve been around. Yeah. Well, you know, part of the reason we built what we built was to solve that exact problem. Just full stop. That’s, that’s the problem.
Stuart Turley [00:15:15] This is not going to be the worst podcast you’ve ever been on. I’ve got to support this.
Jon Brewton [00:15:18] There’s still a ways to go, but now look, in our industry, there’s such a problem with abstraction in sort of the disparate nature of information and how you can connect information. And it’s not easy. Usually sort of in different formats within different systems and you have to have, you know, crazy people that exist within the industry that will come hell or high water. Make sure that I can connect this information in some way. And so you find these people that build these. Readable Excel-based models that integrate pieces of equations from different places so they can try to figure out how to use information from a different place abstracted to the problem that they’re trying to solve now and try to gain some relevancy and some context that will inform how we can solve problems in different ways using different information. You know, that’s that’s an industry problem because of the way that we’re set up. Civitas is a great example of this. So, you know, we’ve been working with Civitas for about a year now, and it started with a couple of different problem statements. It was, hey, we just bought eight different companies in the last two years. And these eight different companies, for the most part, had different reporting systems, had different contracts for the same services, had a different permit. Had different operating requirements, had different commercial agreements, had different parameters that they saw as value-adding for the companies and these assets were being operated in kind of different ways at the end of the day and so Civitas said, can you just help us like put this information together and try to figure out how it’s connected and what it means? Like we understand from a top line perspective what we bought but in a micro perspective we don’t understand really what’s you know which asset is like a key contributing factor to that top line. And which ones aren’t, and which ones are operating inefficiently and which ones are. And so they said, can you just pull this together for us? And we said, yeah, we’ll, we’l do that. So we ended up scanning, I think 386 boxes of legacy data and then harmonizing say about 15 different systems of well repository information together to build a unified and harmonized data model. Now this data model consisted of about 6.6 million different points of data. And it was really geared around, can you just tell me from the perspective of, say, an accountant or an engineer what this well is, what its contributions are, what it means to our profitability, how it aligns to our lease, and how that’s performing. Yeah, absolutely. We built a tool where we could take. Any asset in this entire footprint. I think it’s 1,862 wells, 242 service facilities, a bunch of pipeline infrastructure, a bunch midstream assets for, you know, frac bonds. You know, just kind of everything together, put it all together in one place, in one data model. And we built some tools where we could use AI to say, from the perspective or the persona of an accountant, Can you please tell me what I need to know about this well? Or this facility from the perspective of a drilling engineer. Can you tell me what I need to know about this well?
Stuart Turley [00:18:32] Drilling engineer is going to want to know basically how soon it’s going to come offline or how much water I got to put in, but yet a CEO is going to want say to his investors as he’s sitting there going, but here’s what strip pricing is by the way, if you give me money now, I’m going to be giving you more money back because I give money back to my shareholders. I got the, I’ve already in, I buy it. Where do I sign? I’m buying. I’m I’m ready.
Jon Brewton [00:18:57] Ready. This is cool. So I think that’s problem one. Problem two was, hey, we bought some midstream assets and we’ve never run a midstream business. It doesn’t mean that we don’t know how, but inherently we’d like to understand just as a general imperative, how can we optimize our disposal activities, reduce our total cost exposure and increase our profitability, which gets to your right now at an asset or a management or an executive level, how can I be more productive and more profitable on a day-to-day basis with a new business entity that maybe we don’t understand all that well? And so we built a tool on top of that to understand the entire performance history of these assets, their commercial state, what it meant for how they were operated in the asked how they were being operated, sort of point forward from a planning perspective. We integrated the wells performance history from a production perspective, how fast we drill wells, how fast, we complete wells, what that means for what we put on production, what our production curves look like, what our decline curves look like. Just this holistic understanding of all these really disparate parts of information that exist in every oil company, but are not connected in every Oil company. And so we kind of put them together in a way where we could start to do value chain optimization across the state of Texas and New Mexico on demand. So if we wanted to look at any individual asset, we could say. We want to reduce our total cost by 10 cents per barrel on disposal in this area on a map. How can I do that? And then we built a system that would provide that answer, but not only provide that an answer, provide the simulation background associated to it, all of the information that was used to leverage and answer that question, all the source files that are associated to what information was pulled in to leverage to answer this question, and just created this transparency engine so that we could not only use this thing as it’s most likely designed, again, the imperative we’re working to and the vision that we’re trying to bring to life is that we can build a world where people and machines work better. And make smarter decisions together.
Stuart Turley [00:21:09] This is this is broke on steroids and I’m telling you because as a business case I’m sitting here if I didn’t already know what I was looking for grok would be worthless in so many ways and what you’ve done is you’ve accelerated if I went out and bought a midstream company but yet I wanted to know where the holes are you have just shortened my path to profitability by quarters this is a huge
Jon Brewton [00:21:38] Yeah, it’s really taking what is the interesting way that these tools are built and kind of flipping it on its head in a little bit of a sense and making sure that we can use it for what it’s great for, which is producing answers. But those answers really depend on good quality information and not only just quality information, the connective and defined relationships within that information structure. And if you can do those things, you can start to use these tools at scale and create real value. I mean, the stuff that we’ve done for Civitas to date. We built just the overall database management system so we can take any individual asset on demand from any perspective and say like. What is this? Why is it important? What’s the contribution is to the system? But we also built a saltwater disposal engineering assistant that’s designed to simulate the economic impacts of different saltwater infrastructure designs and disposal methods of produced water. This solution is already identified in planning. So it’s kind of got two applications in strategic planning. How can I be more effective? What are my assumptions? And how and I minimize my total cost exposure on any sort of risk basis to reduce my total. Cost and my total capital outlay before I start a project, that we were able to identify about $10 million in things that you would otherwise not want to do so you could avoid some cost exposure on capital deployment. Then we also built a tactical program where we understand the economics of every well that’s being produced. We understand the economics and the commercial agreements for every disposal asset. We understand the distances between every disposal assets. We understand trucking availability and we understand how to optimize these things to the lowest possible cost. On a footprint basis, so we can circle wells on a map and say, like, I want to reduce my total cost here. We can take it to a lease level. We can say, we want to minimize our total cost exposure here by a certain amount. And that looks to be on a field-by-field basis able to produce between five and six million dollars of cost optimization on an annual basis per field. And so this is about taking ideas and turning them into dollars. And that’s the hard part with AI now. A lot of people, you’ll hear, I love this because this is my business now, but the business that we’re talking about right now is my businesses before and where I work is sort of the intersection of these things. If you talk or listen to the CEOs of companies like OpenAI or Anthropic or anybody else there’s this thing that they say hallucinations are not a bug. They’re a feature because they allow us to think about different things that we otherwise wouldn’t have thought about in the context of what we’re doing. And that’s about the best way to put lipstick on a pig I’ve ever seen in my life. They’re trying to paint what is an inherent flaw and a problem with the systems that they’ve designed as some sort of thing that could create or add value. And the funny thing is in the industries we’re talking about, There’s no way to scale those things meaningfully if they fail at a rate of 50% or 60% or even 20%. And, you know, I think personally, I asked myself this question, what part of the businesses that I used to run or the businesses I used to work in can be wrong 30% of the time and still be profitable?
Stuart Turley [00:25:04] You’re almost a politician at that point, you know, you can be wrong all the time, as long as you get your donations and then steal the election. I don’t whatever you want to phrase that.
Jon Brewton [00:25:15] Yeah.
Stuart Turley [00:25:15] But holy smokes, you know, where we’ve come from since 1977, I said, whatever the number was, I was programming Fortran and a card key punch. How many, how much eutectic material would it take to keep a logging tool from overheating as it’s traveling X number of feet per second. I just totally threw the entire mainframe at OSU into a overload because I miss did my cards. Now, where there is no way we’re even on the same planet now, this is absolutely.
Jon Brewton [00:25:50] Well i mean look when i started in the industry twenty years ago which feels like a long time ago but respectively probably not it you know we were still doing a lot of things by hand. Yeah, we had computers. I was working out on rigsites as a company man on location. I was doing the same things that you were just talking about on a pen or on paper with a pen and a calculator and trying to learn how to sort of do this stuff and excel at scale and trying to figure out how to build these models. And today we’re building these insanely intricate systems.
Stuart Turley [00:26:21] How do we get you to Elon, because Doge was sitting there and the amount of corruption that they’ve been able to discover, what you could do for a government contract, the oil and gas, I can already marketize you. I need to visit with your guys over here because I’ve got all these CEOs and everything I’m excited because I’m already on board. I’m running down the road and I’m sitting here thinking, think about Doge going around and think about what they found and the, and the systems and the errors and the that you could find there, but every single M and a activity you could branch out across, not just energy. I mean, we’re talking scalable of a product set. This is unbelievable.
Jon Brewton [00:27:10] Yeah. Look, so the two domains that we work in are primarily the federal government defense in the Intel space and the oil and gas space.
Stuart Turley [00:27:17] We’re having zero prep work, John. I swear, I did not have a prep meeting on this. This was a question then.
Jon Brewton [00:27:25] No, it is. It’s a great question. And so a lot of what we spend our time. So I spend a lot of time on Capitol Hill, like I spend a lot time with the speaker’s office and the leaders of the AI Congressional Task Force. And I spend a lot of time talking to congressmen and other people that are working. So I’m in D.C. At least once a month and. We spend a lot of our time talking about explainability and explainable AI and government applications. And it starts with trust transparency and traceability. And it, it starts with we really need to understand one that. These systems are inherently flawed, so you have to design different ways to approach problem solving based on those flaws. You know, it’s kind of like if you’re going to engineer anything, you need to understand where the weak points are so you can make sure that you can cover up for it in a design, right? Explainable AI doesn’t just happen. And so you have to reconcile what is sort of a very hard truth about hallucinations and what that means for the benchmarks of how these tools perform and the fundamental flaws with them. And then you need figure out how you can start to harmonize data. One of the biggest problems that the federal government has, stop me if you’ve heard this, associated to our industry, oil and gas, is that they have a lot of different data feeds from a lot of systems that have a whole lot of overlap, but they’re not connected at all and they don’t know how to reconcile it. A lot of what DOGE was going after was trying to figure out smart ways to connect these systems together so that we could gain transparency. That doesn’t happen by accident. And there are sort of traditional data engineering approaches that you can take. But a lot of the discussion that we’ve had with the government is if you package information into the right data model. And so what I mean is, you know, instead of a SQL table or something like that, you put it into a knowledge graph graphs can work on a semantic basis. And so you can convert numbers into words and you can convert words into numbers. And it kind of goes back and forth. Then you can start to embed these things and label these things, which creates a bridge that connects how these large language models work. To what you’re trying to get out of the system. It’s language at the end of the day. And so we spend a lot of time with the federal government talking about high reliability applications for AI and how we can rethink what we do from a database structure, from labeling, tagging, a governance perspective to make sure that we can start to use these tools at scale and start to find some of the fraud, waste and abuse that’s in the system that’s almost certainly. Pervasive in the system, it’s actually as hard to uncover as it is to fix, and it’s because of the core underlying problem, which is none of the stuff’s connected. And it’s really difficult. So we’ve had a lot of discussion with the FBI, with CBP, with some other folks. The VA is another big organization that we’ve got a lot discussions with because they all suffer from the same problems. Too much data, not enough connectivity in that data. And the inability to really sort of use that information to make better decisions. And ultimately we’re trying to help them with that.
Stuart Turley [00:30:23] And not enough people that understand that they’re drowning in information, but yet can’t do anything about it. Wow. What a cool cover. So I’m sorry, I’m fired up over here. I’m already planning about three more podcasts with you because I’m thinking of about all these different things that this could be used for. I’m going to be one of your biggest cheerleaders out here besides Mark, but you know.
Jon Brewton [00:30:48] Yeah. Well, look, you know, I think we’re kind of in this very interesting period of time where when these tools hit the market nearly two years ago, we’re almost there. A lot of people didn’t understand how they worked and didn’t understand how to apply them in different ways. There’s been this learning curve that’s taken place. And I think as people understand more clearly how they work and some of their deficiencies, people will start to figure out ways to use these tools more What we tried to do, and what we do that’s really different than everybody else, is make combining information easy. Make leveraging that information easy, make really interfacing with these tools and the capabilities that they can provide very easy. Our platform, through that approach, reduces time to insights by about 95%. It cuts analytics costs by about 70%. And we can actually secure our database in transit and in rest at a zero trust security and encryption level, which is the highest standard for the DoD. And so that… Makes it easy to start applying to engineering use cases and our industry, the ones near and dear to my heart, the oil and gas industry.
Stuart Turley [00:32:00] We’ve got, not only that we’ve got utilities with grid problems and the Biden administration let back in 420 major grid interconnects that president Trump pulled out and they put back in that can be remotely shut down from China. So you, utilities are like ripe for your service. Holy Batman. This is a huge deal. Well, how do people find you, John?
Jon Brewton [00:32:27] You can find me data2.ai. That’s our website. You can find me at LinkedIn, my handle our name is at John Bruton, J-O-N-B-R-E-W-T-O. And he can find our company page at at data to U.S. We don’t do much outside of that, outside of like, you know, in the story, we don’t have an Instagram account or other things like that. So those are our primary methods for connecting with folks. And obviously you can reach out to Mark Stansberry, who I think everybody in the oil and gas industry knows, and he can get you to us pretty quickly. But yeah, it’s easy to
Stuart Turley [00:33:03] I was really, I would really like to look at this as one of our first podcast in a series with you and Mark, because this is such a gigantic problem right now. We are seeing not only president Trump in, in a whole new world of respect. I get all teary eyed when I see the Saudi prints meet him at the airport. And then when he leaves Saudi Arabia, he leaves with a trillion dollars worth of investment, not only look at that. I knew I liked it. Look at this.
Jon Brewton [00:33:36] Look at this way. Wait, so this is our version of the magnet. We are here to make analytics great again. So
Stuart Turley [00:33:44] analytics great again for my podcast listeners john has got a hat on that says make analytics great. Again, a holy smokes batman.
Jon Brewton [00:33:57] What’s what’s fun about this is President Trump has a couple of these hats and he is in fact signing a couple these hats, and sending them back to us. He’s a fan, at least as we’ve been told. We haven’t had a chance to sit down with them directly, but he kind of understands what we’re up to, I believe.
Stuart Turley [00:34:12] I’ll tell you, that’s on my bucket list is to interview President Trump and also Tyrus off of the Gutfeld show. Tyrus is one of my absolute heroes. That’s on my bucket list. Sorry, Mark, you, I’ve already interviewed you two or three times. That’s off my bucket and then president Trump, but you, you you’re critical for what’s happening because really what happened is happening. Mark and John, we’re seeing a transition from a old economy to a new economy. There’s going to be mergers. There’s going to be acquisitions. And our great energy secretary of Chris Wright has been on the podcast about four times love Chris Wright and he is dead on right. When he said LNG is going to be our biggest export that the US has, but he’s also saying the battle for AI is even is one of the biggest battles we have facing us as a nation right now.
Jon Brewton [00:35:19] It’s like we’re really in the information and capabilities democratization business. We’re trying to make sure that everybody can use these tools and use them effectively. If you don’t prepare your information the right way, you can’t. But we’re trying make that easy for people. The stuff that we’ve done so far is generating a tremendous amount of value. But we know, I think a lot of people think we’re missing data. Like we know for a fact that people aren’t missing data, what you’re missing is the connections in your data that matter. And that’s really what we’re trying to help people with.
Stuart Turley [00:35:50] You’re missing the connections in your data that matters, but you don’t know what you don’t. So how do you know if it matters? This is huge.
Jon Brewton [00:36:00] We can help.
Stuart Turley [00:36:01] I love that. Oh, I had such a great time. Thank you. We’re going to have your stuff in there. We’ve got about two more minutes here. Mark, what are some of your thoughts on this today?
Mark Stansberry [00:36:11] Well, I’m, of course, enthused. That’s why I wanted the two of you to get together. That’s the energy between the two of you, nothing moves without energy. And that’s Stu and John, and then the listeners out there, because this is a world, a new frontier. And that new frontier is so exciting. I can see the years ahead that we can all become active that way. The data centers, all the, in fact, John, if you could just give a brief of what you’re working with at LSU in Louisiana.
Jon Brewton [00:36:37] Oh, nice. Yeah, yeah, absolutely. So we are in a partnership with the Louisiana State University. It’s a broad partnership for AI research and development. We have some strategic go-to-market partnerships with the US, Microsoft, Dell, Nvidia and Neo4j. But on the academic side, there’s no better university in the United States of America, at least for us, when it comes to energy research, and very specifically, cyber operations and government overlap, because they work with the department of the interior, they work with the Department of Energy, they worked with a bunch of folks that are really tied into critical infrastructure and a lot of making sure that we don’t have critical failures there. Is about connecting information. And so we’re doing a lot of stuff with Louisiana State University from the College of Engineering, the College Of Science, and ultimately trying to find ways to bridge gaps between what is high frequency data, sensor data, network data, so we can understand what’s happening on a distributed network infrastructure and try to figure out how we can come up with proactive ways to screen these systems and monitor these systems for real-time failures and problems, and then start to develop engineering tools that we can supplement the industry with. So we’re doing some very interesting stuff at Louisiana State University.
Stuart Turley [00:37:50] Oh, I think that’s fantastic. Louisiana has got its own language and I love Louisiana. I was there years and years ago. And when they, I was working on a wide area network, trying to get it ready to go. And Gator means a wide-area network with 14 gigabit net kind of interconnects. And I’m like, so I have no idea what the guy just told me. I had to have it translated while I was working there, so bring it later.
Jon Brewton [00:38:16] Yeah. Well, luckily I’m fluent, you know, working, working in the industry, working on the Gulf coast. I learned, I learned the language in the shorthand. I think every third person I worked with was a graduate of Louisiana state university. It was either that OSU, OU, or the university of Texas. So that was, that was pretty much every single person I worked with regardless of what function I was in.
Stuart Turley [00:38:36] Well, I do want to thank you for your service and your entire management team. The service, I should have gone in the army, but Jimmy Carter was the president and I did not want to go in the Army when he was there, but my dad retired after 30 years and flew jets all over the place. So I love the military and everything that you guys do and stand for. I wonder, do you go to the range for board meetings?
Stuart Turley [00:39:21] With that, thank you both very much for stopping by the podcast. And I sure appreciate both of you. Thank you very much.
Jon Brewton [00:39:28] Thank you, sir.
The post AI’s Impact on Energy: Optimizing Data and Decision-Making with John Brewton and Mark Stansberry appeared first on Energy News Beat.
“}]]
Energy News Beat