Subscribe
Share
Share
Embed
Welcome to Energy Bytes with John Kalfayan and Bobby Neelon. Your essential guide to the intersection of data and energy. This podcast dives deep into the world of energy, shedding light on how data, AI, and technology are revolutionizing this sector. Each episode equips listeners with insights into the most efficient tools and resources, paving the way for a data-driven future in energy. From technical nuances to broader industry trends, Energy Bytes offers an unparalleled perspective on the evolution of the energy industry. Join us as we decode the algorithms of energy, one byte at a time.
00;00;00;00 - 00;00;28;03
Unknown
Wake up. Welcome. All right, well, welcome to the first, Energy Bytes episode of 2026. I'm Bobby Neal, and I've got my Cassie trusty co-host and Cassie. Yeah, that, I do that to John Callahan, the rad dad. How are we doing? Happy. Happy new year in February. All right. Yeah, I was running the other day. I got saw someone I hadn't seen, and I was like, is it still, like, can I still say, you know, you're appropriate as a two way?
00;00;28;06 - 00;00;47;01
Unknown
But not this is an episode I've been trying to make happen for pretty much since we started time. But I've got, my good friend Hussein Shelf here is the CTO and head of business development for energy and utilities at Amazon Web Services or AWS thing. You got that steel trap, Bobby, is that memory? That's pretty good.
00;00;47;04 - 00;01;05;12
Unknown
Good job, good job. Yeah. Buffer. Yeah. I mean, sand and I go back probably now about ten plus years. At this point, we live in the same neighborhood. And. Sienna, I didn't realize that. Yeah, but, yeah, we're about 5 or 6 houses down from us, so I think ran into you walking the dog, and you and I started nerding out.
00;01;05;12 - 00;01;23;24
Unknown
You were so Chevron, I think, and started nerding out about data analytics. Think. Well, there was an AI and data analytics and Cloudera and Hadoop and all of the things I do. Man, it's been a minute. Yeah. So I've talked about the dude. See he's he's seen it all, folks. Oh this will be a great no gray hairs that I have.
00;01;24;00 - 00;01;41;10
Unknown
Yeah. But I remember that I think this kind of falls in line with our kind of maybe core purpose with the, with this was. I mean, I'm here talking to you, and, there was some you did some coding challenge or something at Chevron, and you, you solved it and I forget what the tool was, but it was like I think it use our code miner RapidMiner.
00;01;41;10 - 00;02;00;24
Unknown
That's what it was. And it was kind of a gooey around like R or whatever. And it's like those kind of things where you pick out. So I never heard of that. And like now it's a tool that you can go try and mess around with. But I mean, I don't even know if it's still around. It's probably a resource by, oh my God, what is the company called?
00;02;00;24 - 00;02;21;02
Unknown
It's pretty popular machine learning. I don't know, I'll remember. Yeah, but it was bought by smaller by a bigger company. Okay. And it's part of their suite now. Yeah, but you won that competition. Didn't you like it within Chevron? So. Yeah, it's pretty crazy. And I was not a machine learning expert. I was not a data scientist, so I think not a statistician.
00;02;21;02 - 00;02;37;02
Unknown
Yeah. So I think we did. I think that this part parlays right into, like, kind of what the conversation where you're trying to not have before we got on here. So we have a meaningful discussion on the podcast. But, it's just like, do you have to be a web developer anymore to make a web app or a software?
00;02;37;07 - 00;02;54;23
Unknown
You know, the, the develop a mobile app. It's just like the the barrier to entry has gotten so much lowered. So much. Yeah. Yeah. Just, you know, curious what you guys are, are seeing a, I mean, it's since we recorded our last one probably wasn't November or December like the advances in I have just been exponential.
00;02;54;23 - 00;03;14;06
Unknown
So, just, anything fun that you've been toying around with lately? Yeah, yeah, we'll talk about all of that. I'm happy to listen. First, thank you so much for having me. I really appreciate it. I know it's been a long time coming and I've been wanting to do this for a while. Just share my experience, my personal experience, my career.
00;03;14;08 - 00;03;35;16
Unknown
Why? I'm here. You know what we're doing today, what we want to do in the future, and and just answer any questions and, you know, help people in their journey. I know how nervous I was when I was just starting out of college, and. Well, can you imagine starting today? Yeah, I was especially exposed to how crazy two a very different, different approach, different prioritization.
00;03;35;16 - 00;03;54;07
Unknown
But just like we were talking earlier, you know, Bobby and I were talking about how you you need to continue to reinvent yourself and just differentiate yourself and see how, you know, what's around you, what you can use to be better. And, and, and provide business value at the end of the day as professionals.
00;03;54;07 - 00;04;12;15
Unknown
Right? No, I completely agree with you. It's the world doesn't change or the world is constantly changing. If you don't change, you're going to get during it, left behind very quickly. Yeah. Actually, these days I feel like, you know, maybe probably around the time that you and I met and I was getting deeper into the text, but it was like, all right, what skills can I learn?
00;04;12;21 - 00;04;31;00
Unknown
Right. That will make me. Well, I mortgage. Yeah. Like, do I need to learn Kafka? Do I need learn, you know, Python or AWS or whatever it is, or Google. Yeah. And now it's more like what ideas I have that are, can I make move the needle and like and now like just your imagination is the limit. Exactly.
00;04;31;00 - 00;04;52;15
Unknown
Like exactly like or maybe some security policies as well. But and you remember like, you remember back in the, you know, the early days we were like, you know, talking about business intelligence and like, you want to democratize business intelligence because you want to give reports, you want to give apps like, you know, Spotfire or power BI and let the business users and know take it to a whole new level.
00;04;52;15 - 00;05;15;04
Unknown
Yeah. Like anyone, it's like we're going to be, like, knows how to write English and eventually every other language, right? Yeah. We'll be able to build an idea and start from scratch and, and deploy and, and have it out in the market for everyone to use. And I think there's pluses and minuses to that. Yeah. But I think as, technologist, the fundamental things have not changed in my opinion.
00;05;15;04 - 00;05;42;11
Unknown
This is a, my personal opinion is that you still need to know security. You still need to understand architecture, you still need to understand scale. And, you know, user experience and performance and all of these things are still the same, whether you are writing the code to do it as I did, you know, ten years ago or, telling the model to do it for you and to see the results and to confirm the results and do that.
00;05;42;11 - 00;06;08;12
Unknown
So, you know, for me, I think that's that's the key thing that customers need to understand is that it's okay to allow your people to experiment and build things and, and and try cool things. Yeah. Cool ideas. As long as you have some sort of a governance in place in the guardrails to protect it from any type of data, leakage or security concerns and typically that's already in place.
00;06;08;12 - 00;06;28;25
Unknown
Yeah, but but even, like, cost concerns, I mean, like, a lot of these things are built on demand now. Right. Like how long is my server on. That's what I'm getting charged for. Right. And it's like if yeah, you can let I write some of this, but if it's not optimizing it and it's just slop and all of a sudden, like now you're talking about scale and it's getting hammered, you know, with requests or it's just a big job and now it runs forever.
00;06;28;25 - 00;06;58;24
Unknown
Or someone who ends up just like on some infinite loop and it just running forever on a huge machine that it's that I spun up for you. Yeah, yeah, yeah. I was like, wait, how do I get a 50,000 hour bill yesterday? Oh, I saw a, a I think it was on Twitter, but I saw this post this guy made and somehow his clawed bots had, you know, the whole premise of that is you can text it through WhatsApp or slack, some kind of chat interface, and then they reply to you, but they're only supposed to act off of your responses.
00;06;58;24 - 00;07;24;19
Unknown
Well, somehow his got messed up and they were just talking to each other. And, it believed it was just taking responses from each bot. And so overnight he woke up to just this giant bill. He maxed out all his tokens and all this stuff. He had to just turn it off. But it's stuff like that that it's like, yeah, I mean, everyone that's ever touched a cloud provider has turned on a server and forgotten that they turned it on and they got the bill and then like, shit, where where did that come from?
00;07;24;19 - 00;07;53;17
Unknown
Yeah. No, I mean, and again, this is because it's the technology is moving so fast. The level of experimentation and unknown is just increasing tremendously. And it's just it goes back to the fundamental. Yeah. Like you, you can put a limit on how much you want to spend, right. Like, you know, it's just a basic setup stuff for a budget of $1,100 or whatever you want, you can put limits within the code you can have the, you know, the guardrails in place, use and file.
00;07;53;18 - 00;08;11;28
Unknown
Right? Yeah. You know, and so I think those fundamentals will not change. And that's why when I talk to people that are worried about their jobs and worried about, you know, how it's going, this is going to impact their job. To me as a developer at heart. I've started that way, and I built my kind of career. And my reputation at the beginning was around that.
00;08;11;28 - 00;08;39;18
Unknown
And understanding the fundamental principles of system architecture is that those are not changed. And now you simply amplifying what you can do. You can run, you can now we all know how difficult it is to work in teams, right? Everyone has opinions, everyone has their own ideas, and sometimes it's conflicting. And you're lucky if you're with a, well, kind of, simulated and well, assimilated team.
00;08;39;21 - 00;09;06;22
Unknown
But now you don't need to. You have agents, you can tell them what to do. Yeah, well, that's what came to mind when you mentioned, like, front end and stuff as it's like, because the agents have gotten so good and they're trained on certain, you know, tasks and tasks, styles, techniques, whatever, like even some of the front end stuff that it automatically built, like I built a a document labeling Streamlit app.
00;09;06;22 - 00;09;30;19
Unknown
Right. And it spun it up and I didn't tell it like much. It's like a on a view the PDF I want to know like the framework I don't care about anything. Right. And it did a pretty good job. Like I didn't have to tweak it very much and it was just the first time. Right. And so like it's going to be very I'm fascinated to see how all of that, because I liken it to like when we were growing up, you had to defrag your hard drive.
00;09;30;22 - 00;10;00;12
Unknown
Yeah. Like you had to troubleshoot your computer. I love the visualization. Right. Moving those stupid dots. But like now that just no one does that is it happens in the background. And so there's there's always this like technology curve of it gets more complicated. But they also abstract out. Yeah. Complexity the complexity. So that now it's like kids that grew up with computers nowadays have no concept of how they really work, because they didn't have to debug all this stuff and figure it out.
00;10;00;14 - 00;10;16;15
Unknown
And I see that kind of potentially happening. And the coding. Yeah. At least. Well, all of us here have had to touch and write some of our own HTML and all that. And you're like, all right, I understand what it's doing. But like there's other people and it may mean sometimes it matters, sometimes it doesn't. Some people it's like, I don't care what it does under the hood.
00;10;16;15 - 00;10;37;04
Unknown
Does it look the way I want it to and it works. But but again, one thing I thought of too, just cause it said like, the nail art in the first shot, you know. But I think another coding principle or principle that comes like a style commit early, commit often, like, you know, you're you should be using GIFs and like when you when it gets to a point that you you want to be able to come back to that you better you better commit that bad boy.
00;10;37;09 - 00;10;56;01
Unknown
Yeah. Because you might want to roll back to it after it jacks up the next up. You know, my my experience is, you know, a lot like these models are they still have, you know, hallucination and they still have their own way of doing things that could potentially break things for you that you've already, you know, finished.
00;10;56;01 - 00;11;21;12
Unknown
Right. So I see that all the time, you know, with some of the stuff that I'm doing on my own, like playing around with, some of the developers, the vibe, coding ideas, etc. and you have to provide some guidance. You have to write some principles, you have to give it very strict guidelines of do not touch this or do this and make sure you confirm with me, because sometimes it just creates its own plan.
00;11;21;12 - 00;11;41;15
Unknown
Yeah. And if you don't review the plan really well, whether it's a spec like in, you know, Cairo or it's anti-gravity or cloud, code, etc., and you, you have to give direction. And again, to me that's not different than telling your developer as a business analyst. Absolutely. Or as a product manager, like, you have to give guidance.
00;11;41;18 - 00;12;04;09
Unknown
And you used to write all these documents, right? Now you got, you know, you're just chatting with the model that's writing it and it's, it's phenomenal the amount of acceleration. Yeah. And the amount of, productivity that that generates. And and we're seeing it like for me, the difficult part is we, I got to stay within the, the, the the framework or the scope of an enterprise.
00;12;04;09 - 00;12;24;07
Unknown
Right. So there's a lot more a lot less freedom. And what people can do within an enterprise that, you know, is conscious about security and, and data privacy, etc.. And so you have to reaffirm that and you have to have a kind of a data strategy or a data foundation strategy in order for you to enable these people.
00;12;24;07 - 00;12;45;15
Unknown
Otherwise it is a wild West. Yeah, yeah. You could become pretty dirty, and chaotic. Do you see that somebody on the DoD got flagged for trying to upload documents to. Yes. Yes. And like. And he was like one of the guys making the decisions for what models to use and what it's like. I was just like, oh my God, what are we doing?
00;12;45;17 - 00;13;04;06
Unknown
But I mean, we see that all the time where it's like the, the company, the operator that we're talking to you, they don't have access to any of these tools. But then they're like, yeah, the CEO was trying to get it to analyze some, you know, report financial. You're like, what? What's happening? It's yeah. Yeah, it's it's still a wild West.
00;13;04;09 - 00;13;19;28
Unknown
I think, you know, it just in general, like, and I've always kind of hated the term data governance, but I think it's bigger than data governance. But it's just governance governance in general. And like even as funny because like what over the last five, ten years, it's ESG has been a big thing. And all everyone thought about was the environmental side of for ESG.
00;13;19;28 - 00;13;40;07
Unknown
But there's this whole US governance in general. And it's like you've always needed it, but now it's like more important than ever. I think, you know, just the absolute, the guardrails. But it's funny because it kind of moves up and down the scale like you have your own governance. So even when you're within your chats with cloud code or whatever, where, or you've got a curse or file or all these like, these rules are like you limit yourself to these.
00;13;40;07 - 00;14;07;21
Unknown
Yeah. You know. Yeah. Instructions are this, you know, this vector go for. Yeah. Yeah, exactly. But then it goes all the way up to the corporate level or, you know, but you know, one. Yeah. No, that's, that's, it's just, it's it's we I feel like we were already trying to catch up on the data governance side, and now it's probably even more amplified because you can download these things pretty easily.
00;14;07;23 - 00;14;26;16
Unknown
It's it's for me, like, you know, it's you can't run away from it, right? Like, as a, as someone, as an executive or a technologist in a company that at the end of the day, you have to protect their own sovereignty. And I mean, data and security and privacy and all of these things. But at the same time, you want to open up the innovation.
00;14;26;22 - 00;14;51;07
Unknown
Yeah, you can still it's not, you know, at one or the other, you can still build the foundational pieces. You can still partner with your software provider or your cloud provider to build the environments that allows your people to do so. Yeah. And then it's just a matter of data access, right? We've we've done data access, principles and standard operating procedures center for decades.
00;14;51;10 - 00;15;15;26
Unknown
Right. Those systems of record are there as long as you and make it, secure enough to allow people to experiment and try different things within the confines of what they have access to, it doesn't change much. Yeah. Whether you're doing it with an Excel spreadsheet, right, or you're doing it with, vibe coding, whatever, pick your favorite application right when the user's going to do it.
00;15;16;01 - 00;15;31;13
Unknown
Yeah. Most like especially the engineers in our industry. Right. Like we just solve problems. And so the more barriers you put in my way, the one that you want them to solve the problem problems through the proper governance checks. Right, right. So you have to enable because like again this is not new. This is shadow it, this is like the old slightly.
00;15;31;13 - 00;15;48;08
Unknown
They're going to find a way to do some bizarre thing through Spotfire because you didn't give them the right access. But if you just give them better access to the database or whatever, they probably could have done it directly and done it the quote unquote right way. But they don't have to. They're they're going to find a way, don't you know, like the locally and, and then it's even more of a security issue.
00;15;48;09 - 00;16;08;24
Unknown
That's why I like going back to my comment earlier. You know, I think before we started is I think people are going to always find the tool that they like to use to solve a problem, and you will never be able to stop them. It's making sure that those tools are vetted and and just don't, you know, share privacy and data security with something else that you're not aware of.
00;16;08;24 - 00;16;26;20
Unknown
And, and that, you know, brings, you know, the conversation in the comment earlier about, you know, why why did I choose RapidMiner or Knime? At the time it was like an open source. Yeah, drag and drop machine learning workflow generator. Right. And I was because I hated writing. Yeah, hated writing Python. Like I don't want to write code.
00;16;26;23 - 00;16;48;24
Unknown
I'm done writing code. I wrote millions and millions of pl SQL code when I was back in Chevron. In my mind that I hated. Right. What could get me the result faster? Right? So I can show the results to the business partner? Yeah. And that acceleration is exponential. Now, like within minutes you can show an application. Yeah, right.
00;16;48;26 - 00;17;16;17
Unknown
Within minutes you can create, mock, you know, pages and sites and applications. It took me two weeks to build an application that is geo based application personally. Right. Playing around with some of those codes, some of those tools, within 2 or 3 weeks, it was already in the App Store being reviewed. I would have never thought that, you know, fully functioning, you know, multiple users interacting with each other, simulated users like.
00;17;16;21 - 00;17;34;05
Unknown
And all I had to do is just tell it, do this, do that. Yeah. Fix this, fix that, copy this error. And and it was phenomenal. This is stuff. This stuff is powerful I'm very excited. And then the agents world. Yeah becomes the other frontier of like okay I'm you know I want you to build a marketing I'm selling Bobby.
00;17;34;05 - 00;17;54;06
Unknown
Only I wanted it to build a marketing plan. But when the review is finished and now people can download this app. Okay, give me a marketing plan that has $0 budget. Right. I want you to do everything in. I. Oh, videos, pictures, posts and it it. Now I have like ten different videos that are all awesome to look at.
00;17;54;06 - 00;18;16;23
Unknown
It's like 4K. No. Yeah. And again and like that you can have that deploy it to all the major, you know the platforms just API integrations. Yeah. And finally you could even get down to the point to find the best time to post on this platform, right? Like, I mean, there's there's a lot to be building, like do market research, find the top five trending topics in this subcategory, generate posts.
00;18;16;24 - 00;18;37;08
Unknown
And like. I saw a post on LinkedIn before, earlier today and it was this guy using cursor and he's got like assets folder. And then you know, a scripts folder and a videos folder and he's doing a cursor and it's like, yeah, I Colin is, is asking us to put out some content. And I've literally this morning asked cloud Code.
00;18;37;08 - 00;18;58;28
Unknown
I was like, look through all my cloud MD files from last six months and give me five blog topics to that I can write on there like, holy shit. Okay, yeah. And now it's gonna write it for me. It's like, yeah, it's really so, so like and I want to bring it like back into the enterprise because I think we, we, we deal with a lot of companies that especially in our industry that are hesitant, a lot of times to trust them with it.
00;18;59;01 - 00;19;25;09
Unknown
And I get it right. And that's where you start looking into chain of thought. You start looking at this black box. What is it doing? What is it thinking? How does it come to this conclusion, conclusion in this result? And you start kind of building that trust with people. And so some of the things that we're doing around foundation models, for example, like you're both trying to do domain specific foundation model beyond just the language or, you know, internet data that it was trained on, right?
00;19;25;09 - 00;19;48;15
Unknown
Yeah. And we're seeing some phenomenal results, some partners that are doing it. We are doing it. Operators are. Yeah. That's what they're doing. Adopting. Yeah. And look at whether it's a well model or a seismic model or reservoir model and, you know, some of the stuff like it's surprisingly, well, you know, the results are really, really surprisingly well.
00;19;48;17 - 00;20;09;13
Unknown
And so I think we're only scratching the surface when it comes to this in the context of energy and oil and gas, where we, you know, have worked and, and, been involved in, so I encourage people to just test and try and play and then worry about the, the, the governance in the piece after as long as you don't deploy the production.
00;20;09;19 - 00;20;26;24
Unknown
Yeah. Just run with it like this. You don't do that with any application. So then it comes down to the, definition of production. Right. Because Higgs. Yeah, I got it was like with Diddy long on and like, yeah, he had his whole Twitter thing. But basically the idea was like you would it was asking why, why you couldn't push this Python code to production.
00;20;26;24 - 00;20;44;24
Unknown
It's like, well, blah blah blah. He's like, well, by that definition, you've got a thousand Excel sheets in production. Yeah, absolutely. Yeah. What are you drawing the line. Yeah. Where do you draw the line? So I'm interested. And again, we don't have to make this specific to us. And, you know, like, you obviously work there, but like, I think of as a cloud platform in general.
00;20;44;24 - 00;21;10;00
Unknown
Well, there's the major, but like, there's this very interesting value chain that you like, again, you go all the way down and say, what, few months ago, AWS with Chevron and GE are trying to do some data center stuff out in West Texas. So you're you guys have your hands on that stuff, right? Like so. So we're so, you know, from an energy standpoint, as a business unit and I'll maybe tell you a little bit about, what we do and then what I do personally, what we're involved in.
00;21;10;00 - 00;21;32;19
Unknown
So our team is, customer focused. We focus on driving innovation and digital transformation and cloud migration. Right. We're set up in a way that we can support all the IT workloads. I think of a SAP back office, all of that, all the way to line of business transformation. So my focus is primarily on the line of business transformation.
00;21;32;19 - 00;22;09;11
Unknown
That's where I came from. A lot of the work that I did at Chevron was, was based on the work in pipeline or, you know, production optimization, water, CO2 flood, blah, blah, blah, blah, all the workloads and then ended up, doing some works in unconventional. That's when kind of we met around the time where I was doing, I was looking at the use case was funny enough is like, how can we figure out, well, control events based on data that was captured unstructured, like a data that was captured in drilling reports and in daily reports that were sitting in comments section that nobody think about.
00;22;09;19 - 00;22;30;06
Unknown
Okay, so we're all thinking about measurements. Yeah, right. Pressure, temperature, etc.. And and so I wrote a program just to simply look at those comments and build a pattern recognition of, you know, there was a control event this date at this time. Here's the comments before. Did it infer some of it? Yeah. And for the most part, it did.
00;22;30;08 - 00;22;56;14
Unknown
You know, because people were talking about what was going on. Right. And that's how we think about it. And so, so those are the type of applications that I was working on in, in, in Chevron and then transitioned to Microsoft, where, I was asked to kind of help drive and build the industry, for energy strategy on what we do with customers, what we do with partners, what we do with technology and services and products.
00;22;56;21 - 00;23;31;10
Unknown
And that's kind of and what I do today is a continuation of that is how do we, grow the business, how do we drive customer success, how do we change the way people are doing things? Right. And that includes conversations about how do we source energy? Yeah, what kind of partnerships we need from our standpoint to be able to drive the scale that is expected from our customers, what services we need to put where, what products are working or not working, you know, what are the gaps in the engineering services that we build that could not meet some of the requirements of energy?
00;23;31;12 - 00;23;56;18
Unknown
For example, seven years ago, one of the main challenges was how do you run high performance computing, for seismic imaging, processing, reservoir simulation, etc.. So HPC workloads in the cloud and the cloud was infinite, right? This is the is the you know, the claim. Right. And and at the time it was not for HPC and a certain extent it is not yet sure even now.
00;23;56;20 - 00;24;25;29
Unknown
Right. It's much, much better. We can run supercomputers now for the AI stuff, but it's focused on AI. But there's there's a ton of demand for that of demand for it. Right? So it's, you know, capacity is scarce, etc.. And so my job, 50% of my job was to try to fix the platforms that I was running on, whether it's compute, storage, networking, software that had to do with, HPC workloads to be able to run the large customers like Chevron or Schlumberger or SLB or others.
00;24;25;29 - 00;24;46;03
Unknown
Right. I'm very proud now in, in DGS or in AWS with the partnership that we secured with TGS, after the merger with PGS, they're one of the biggest seismic acquisition companies. They run a significant amount of HPC. We're running it now in each and in AWS. We've been doing it for the last 4 to 6 months.
00;24;46;03 - 00;25;05;13
Unknown
They have the biggest multi client data storage in the world. And so think about now what happens from a GDI standpoint in a large domain foundational model when it comes to this data and how we can use it for our customers to change the way they do interpretation or imaging or whatever, you know, workload that they're looking at.
00;25;05;15 - 00;25;30;07
Unknown
That's kind of the broad brush of what we do. And then I can go into specifics if you want to, some of the things or just talk about the, the, the platform and the technology. Yeah. I mean, I think we can get into all that, I guess, where I want to go that is just just that I think when you think of a platform like AWS or Azure, GCP like they're at every level of of it, like literally down to like you, like you, you need the energy.
00;25;30;07 - 00;25;49;05
Unknown
You need the data centers to provide this service. Right. But then, you've got, say, folks like Snowflake and Databricks that sit on top of you, that need your services like this, that whole with that concept of like digital five of them. Right? Yeah. But like you guys literally get down almost to the atom level. Yeah, yeah. You know, and then you go to the electron.
00;25;49;05 - 00;26;21;01
Unknown
Right. You know, right now we're starting at the electron. And don't forget AWS is one of the biggest custom silicon, maker in the world. We build our own arm based, graviton and Trainium and inferentia like, hardware. Yeah. You know, a significant amount. I don't remember the exact number, so my apologies, but a significant amount of our services that run in AWS run on our own hardware that's crazy and can't control the energy efficiency, the performance, the cost.
00;26;21;04 - 00;26;42;08
Unknown
And it's been phenomenal. And of course, we support and talent AMD and Nvidia and everything in the world. Right. And we partner with the Databricks and stuff, etc.. Right. And we want our products to perform at the level that our customers expected as well as our partners. Yeah, a huge part of my job is to make sure I, Salvi, runs well on AWS.
00;26;42;08 - 00;27;12;20
Unknown
Yeah. Back when I was at Microsoft, Halliburton, Weatherford, Baker, etc.. So those major partnerships and again my focus is line of business. Right? So my focus is those big partners and small ones and I work significantly with, startups, for example, because there's always a cool thing that's happened out there that either we see and we want to bring to our customers or our customers are invested in it and see it and bring it to the market.
00;27;12;20 - 00;27;49;18
Unknown
One great example was the partnership we did, I started I started back in Microsoft and also here AWS is with bluer. Okay. Blue. Where was you know, they've done phenomenal work around optimizing seismic data, seismic storage, but also interpretation of that data, right. And interpretation of 3D models. So they had, I think, one of the first to build and deploy the reinforcement learning model, where a user can literally on the fly, you know, do all of the outlines and the interpretation of a, of a slice of a cube, and then it will infer that, for the rest of the 3D oh, wow.
00;27;49;18 - 00;28;11;05
Unknown
Cube, and then it will come back to you and then you can scratch and improve. So it was a back and forth. It was phenomenal. Right. And so, you know, as you said, the stack is from the electron all the way up to the software. Yeah. And the question is, you know, what is the workload? What is the problem you're trying to solve with the customer?
00;28;11;07 - 00;28;27;17
Unknown
And how do we bring the right people to do it? And so sometimes you find me all the way down to the hardware. Yeah. Or to, oh, we gotta, you know, figure out how we're partnering with customers to bring, you know, renewable or green energy into our data center as part of the partnership that we have with them, etc..
00;28;27;25 - 00;28;44;02
Unknown
So it's all over the place and, and it's fun. And then it's like, second of all, because, you know, you help your energy companies do what they do better. And now it comes back through, you know, to you all as well. So yeah, that's just kind of a wild mind if, if you. Yeah. No I mean all of them.
00;28;44;02 - 00;29;06;10
Unknown
Right. Like, you know, even now the most at least I know AWS and Azure both have pretty much all of the models able to be hosted in their studio. Yeah. Says Gemini. Right. And it's like I think it's that's one of the more interesting things is everyone's kind of converging into the same areas, at least on the AI side.
00;29;06;10 - 00;29;29;11
Unknown
Right. Like you've got a studio, you've got to have storage, you've got to have compute. And so it's going to be interesting over the long run to see where the different like, yeah, how what the differentiators are and how that impacts or doesn't impact the market and stuff like that, I think from the AI side. But I've got so many questions I'm trying to find to organize them a little bit.
00;29;29;11 - 00;29;48;12
Unknown
What just currently, what are some of your the more interesting, either ideas or tools you've played around with that, that you find kind of interesting. And so so I've been, I've been playing around with like a lot of them and it's, it's part of for me personally, it's part of continuous learning a journey. It's just impossible to keep track of everything.
00;29;48;14 - 00;30;09;07
Unknown
Right. And so looking at, you know, from an idea standpoint, I, you know, tested of course, our, our own gyro, which has been phenomenal, very VSCode like, which is what I'm used to, I've spent decades in and, you know, visual Studio and and so and is that separate or do you have to go through AWS to get to that IDE or is it available?
00;30;09;11 - 00;30;28;16
Unknown
No, I think it's available now. I'm okay. Okay. Download. Oh it's like it's an idea. It's, it's it's Yeah. And then cloud nine or something. No no no it's not. And then I tested Curser and I tested, cloud code and, and anti-gravity and, and I think they're all I would say I've always been a believer of an 80, 20.
00;30;28;16 - 00;30;46;12
Unknown
Like, they're all pretty much the same. Yeah. You. It's just a matter of what you enjoy. Yeah. And what you feel comfortable use. Right. So I started with Kira, and we have a very unique thing about it with spark. Right, right. It's based on how we think about building services structured way like we talked about. You need the guardrails.
00;30;46;12 - 00;31;11;13
Unknown
You need the requirements. You need the limits in there and the direction. It excels at that in my opinion. But I also like, you know, you know, the skills in, in cloud code, right? Or I like the anti-gravity, you know, multi-agent manager because you can actually organize it separately. And I've been using all of them and, you know, stick with one for a while and then stop and go there and say, oh, they have a new feature and that's it.
00;31;11;15 - 00;31;29;17
Unknown
So at the moment, I think there's a lot of that, but I think eventually is going to be, 1 to 3. Yeah. And again, it's the same thing, you know, with Java developer. Net you were you were a Visual Studio guy, you know, and eventually people will just get used in for me, it's just a tool point is absolutely.
00;31;29;21 - 00;31;53;09
Unknown
Can I use different models? That's where the differentiation is. And so whether it's Nova from AWS or, you know, chat or OpenAI or Gemini or Claude, you know, the ability for me to access and see. Because if you look at the way they execute, they are there are differences. There are differences in how they organize their thought or their reasoning.
00;31;53;11 - 00;32;11;08
Unknown
And I've I've liked how Claude, especially opus was, was doing a lot of work versus Gemini three when it came out was phenomenal and I liked it and I like playing with it, as well. And then I test Nova on the side as well, just to see the continuous improvement that we have in AWS. And so plus I got unlimited tokens.
00;32;11;08 - 00;32;36;15
Unknown
Yeah. Right. So, so I think for me it's, it's no longer about the, the decisions that you have to make on the technical stack. Right. It's now about the idea. Yeah. And how do you get to the result of the idea, which is as a developer, it's a freedom, I completely agree. Yeah. You started in the infrastructure stack.
00;32;36;18 - 00;32;54;12
Unknown
Cloud kind of brought that freedom of who cares? Serverless. Yeah. I just want to run code and see what it does and get a response. Yeah, right. I don't care where it's running, how big, how small. I can tune that with a slider. Now you're getting to the point where, I just want to build an idea, right, right.
00;32;54;12 - 00;33;36;17
Unknown
And I don't care how and where and all of that. I think is important to, to start. So it's just a mindset of how do you see that impacting especially industries like, you know, energy? As far as like, you know, traditionally very slow, very hard to, to, to take an idea to production, so to speak. Because I think that is one of the biggest like hurdles from an industry perspective is, you know, we either have these pilots that on a small scale or successful and they can't figure out how to scale it up, or you just have a, you know, dozens of different individual scripts that the engineers wrote to fix their
00;33;36;17 - 00;33;56;20
Unknown
problems. But yeah, no one's sharing that across the company, even though if they did, yeah, more people would use it. It would get better over time and those sort of things. But I think one of the with, you know, we have the PhD business model and one of the most, the wildest things is how quickly we can go from scoping with a client to a prototype that they are testing.
00;33;56;21 - 00;34;16;29
Unknown
Right. And then because that to me, one of the biggest like, well, the hardest parts of any kind of software or app development is that feedback loop with the client. And and then it normally it takes you a very long time to get to the point where you have an alpha or beta that they can then start, and you might have totally missed the mark with that by that point.
00;34;17;01 - 00;34;38;14
Unknown
Now it's like, oh, you wanted it. This way. Okay. Hold on wait, wait. Now it's just like the the, you know, like the ability to get to a certain thing and think how that impacts like scrum and all this kind of like for me it's, it's it's the freedom it's delivered. So to the from an enterprise standpoint, I've seen the best and I've seen the worst.
00;34;38;17 - 00;35;01;29
Unknown
And unfortunately it has to start with a cultural change unless there's a cultural change in typically and I hate to say this doesn't really just come from the CIO. Right. It comes from the CEOs. Yeah. Because most of the companies that we work with, right, are energy companies. That is their core business, not I.T, unfortunately, and a lot of IT people I hate to say this, you know, some of them are my friends.
00;35;01;29 - 00;35;17;23
Unknown
Yeah. You know, don't think that way. And I'm like, no, this is this. You are simply supporting the business and you have to and you have to understand what they want and what they do, and you help them there enable them. You want to be in tech. They'll be in tech, right? For tech. But you know what? I left the industry, came to tech.
00;35;17;29 - 00;35;42;04
Unknown
And I'm in the service of the industry. Yeah. So yeah you have to it's hard to get away from it. You have to solve the problem like there's a business problem. I get it, there's a technology, but it's like I think it's tough to it. There's kind of a dichotomy because like, again, I think if I was a small piggybacked operator, I'd be like, give all my engineers and smart people just give them access to these tools and say, go wild, like build some things that do stuff that we couldn't do before.
00;35;42;05 - 00;36;01;08
Unknown
And you get a lot more freedom there. But then if I look at, you know, I feel like I've seen stuff from like, say, Justin Low, Chevron, like they've created these little innovation centers within their companies and like, they're moving fast and breaking things within the confines of what they've got there. I mean, like Alex at Oxy, I mean, like, I mean, there's some really cool things happening inside the walls of some very big companies.
00;36;01;10 - 00;36;15;07
Unknown
But they have to enable, I mean, like, there's, there's more governance or they've been given a little sandbox, but this but let's go back to the Chevron example. I love that because, you know, it's my dear and heart to me. And I was part of that creation. I at least a small part. Justin, don't get mad at me.
00;36;15;09 - 00;36;30;28
Unknown
But I was in the conversation. I just put it this way, and the and it was all driven by Mike saying, you know this, we are an AI company. We're going to drive AI and technology company, etc.. Yeah. So it came from the CEO down to the mandates of, you know, here's how we're going to do it.
00;36;30;28 - 00;36;47;04
Unknown
The partnership with Microsoft that they had were all and there was no debating. Yeah, right. It was it was something that they have to do. They work with them. They sold their data centers. It was a it was a lot. But that was how they were able to start South Central. That was the start. Right. Well there was South Central or yeah.
00;36;47;04 - 00;37;04;19
Unknown
South central. I thought they bought the data center. I thought they bossed the servers in San Antonio from Chevron didn't they. No, no, no. So there was a region. We extended it. Okay. To the because there was literally across the street. Okay. Right. The joke was we'll just plug in the network. Yeah. Right. Or, you know, through the window at five.
00;37;04;27 - 00;37;27;26
Unknown
Yeah. And I can't even say it, but I think in some instances we had to do that for a certain thing that I'll talk about, maybe later. But to the point is, there was a cultural change and an empowerment for someone like Justin and Marjorie at the time. Right to say. And the CIO, of course. Bill, at the time to say we need to do something different.
00;37;27;28 - 00;37;48;09
Unknown
Yeah. Go take advantage of partnering with one of the biggest technology companies in the world, right, to do something different. And how do you set up the governance and how do you set up the security? And then, you know, and it all starts with these framework, whether it's MLOps or DevOps, right. To create a guarded, funnel. Yeah.
00;37;48;09 - 00;38;10;22
Unknown
And then you work with your business partners. And this is, I think, where someone like Steve, Bowman was phenomenal in his, partnership with the technology companies, with Justin and others. And he's on the drilling side where he helped champion a lot of these initiatives. Right. And he brought the business in it together to work on something that is going to be transformational.
00;38;10;24 - 00;38;30;00
Unknown
And all the governance is through them. All the you know, the prioritization. Right? Because you'll get thousands of ideas. How do you pick the right. And that's a big deal before you start coding. Right. And so these these things change from a cultural standpoint. And I've seen, you know, a lot of work that's going on around oxy who's doing a great job with us.
00;38;30;02 - 00;38;58;12
Unknown
Right. Do they're a big partner of AWS. With the gentleman Patrick. Okay. Yeah. I don't know him. I know we like yeah, I work with Alex Lak, at Conoco, and he's been a big part of their. I like, you know, enablement team or. Yeah. Yeah, yeah. So. So, their new chief. Yeah, I, I can't believe I met him multiple times and I apologies, I, I'm just escapes me the names he sees me, and he's, you know, he's done a phenomenal job where we're deploying things left and right.
00;38;58;14 - 00;39;19;21
Unknown
They're enjoying it. There's a return on investment associated with it from a business value. So you can quantify, you know, this pilot that became a production? And I think there is no excuse to not do it the right way. Yeah. Because the frameworks and the technology is there. Yeah. You just have to build the right.
00;39;19;21 - 00;39;42;21
Unknown
Yeah, you have to be intentional. We have to be intentional on how you enable your engineers. Yeah. I think one of the things you you said as well as like that, I think a lot of people get hung up on is you you didn't say scoping, but you were talking about scoping. You know, it's like the problem with AI is that you've been led to believe it can do everything for everyone, even.
00;39;42;23 - 00;40;06;28
Unknown
And there is a world where, yes, but you have to build to that point. It doesn't just you don't just give if I give it to the AI is my favorite tagline of, 2025. 70% of what we do is typical traditional machine learning, right? And data right. Like I'm data engineer. Yes. Like I think that's one of the, the more, surprising things that the average person just doesn't know is like a rag.
00;40;06;28 - 00;40;29;21
Unknown
It's just a giant data search problem. It's not it's not, you know, some crazy, sexy Lem, right thing, but it's intentionally done that way. And so. But the scoping piece. Right. Like I and I've witnessed this well before I was ever here doing AI stuff, but operators or people just generally have a tendency to want to go solve the biggest hairiest problem first.
00;40;29;23 - 00;40;47;11
Unknown
And it's like those take time and they have to have a lot of buy in and like, you have to like there's so many unknowns and all this stuff. And it's like one of the things that we've gotten really good at that I encourage other people to do the same at is when you're sitting down, just try and have you got a giant list of problems.
00;40;47;14 - 00;41;07;11
Unknown
It's like, find one that is a low hanging fruit. That's a quick, easy win. And then build trust. Build. Yes. Exactly right. Like your manager isn't going to trust you to go do this giant thing if you've never done even a small one before. 100%. And so and you also don't know what you don't know about the big hairy thing until you do a bunch of small wins and then you win.
00;41;07;12 - 00;41;28;21
Unknown
The process begets success. Exactly. And often those big hairy things like they there's multiple steps to get there anyways. Like right. Exactly like it should have been. It wasn't scoped down enough. Yeah. Like I mean I've got a client now and it's like they wanted this asset scorecard thing. It's like, why don't we build. But it requires multiple data sources to get there.
00;41;28;21 - 00;41;46;27
Unknown
And like, why aren't we building each piece of this? Like, they're gonna are all gonna eat up their own data model, right? To get there anyway, it's like was build the individual pieces and then now he just picked the right the pieces of that that you want. But it's like they want to do you want to jump to this thing which is like a, like a down the road rat type thing.
00;41;46;28 - 00;42;06;05
Unknown
You know, you just build roam in a day rather than just like I agree. And you know, I've seen like, oh man, remind me a lot of like big data foundation programs that fail. Definitely because of that. Yeah. Because it was it was a bunch of architects that were thinking academically. Oh yeah. Wanted to perfect answer before you can.
00;42;06;07 - 00;42;26;03
Unknown
And you know one of the things that I you know I've been lucky and knock on wood is I've always been a practical architect. Yeah. Meaning like there is a problem with trying to solve what is the fastest path to solve it. And then I can figure out the rest where you build momentum and. Right, you need to build trust and momentum and show results.
00;42;26;06 - 00;42;42;12
Unknown
It doesn't have to be 100%. Yeah, nobody expects perfect, right? It never was. It never was. Yeah. Right. And so my favorite argument this is like what do you mean? It only gets me 80% of the answer and it's like 80% is a lot bigger than zero, which is what it currently is right now. Don't get me wrong.
00;42;42;12 - 00;43;00;27
Unknown
There are things that have, you know, implications on security or safety that you have to, you know, think about airlines and even even the pipeline, you know, etc. for us in the industry, like those are important things that you want that rigor. Yeah. And I can tell you 90% of the time they're not 100% right. They're 99.999. Right.
00;43;00;29 - 00;43;19;22
Unknown
So it's a combination of how many times you can you can be okay with. But but the important thing is, is to build that trust in the show, the results. And now as we said in the technology like you can with the new technology, you can really, you know, from an idea to a prototype, it's minutes. Yeah.
00;43;19;22 - 00;43;36;22
Unknown
That's amazing. That's so cool. I think you get on to like, just like the average person. Like I still I think we forget like, we, we sit in our circles of people that are like, you know, and we're on the, you know, very far end of the bell curve. And it's like, I still talk to my parents or my wife, or it is like, don't you shouldn't be doing that.
00;43;36;22 - 00;43;57;20
Unknown
Like it's like, stop doing. Yeah. Like, well, here's a funny one for you from a from a professional setting. I was asked to present at a group of, private equity, firm, and they invited all the CEOs and CTOs of their companies that they, they, you know, they're, whatever they call them, right? Investment companies.
00;43;57;23 - 00;44;16;23
Unknown
And they were presenting all the success stories. And I'm saying this was, you know, not like six months ago, maybe eight months ago. And I'm in New York, I'm meeting with them and I'm a, you know, guest speaker. I'm talking about what we're doing all the crazy cool stuff. And after after they were presenting and I'm I'm not joking.
00;44;16;26 - 00;44;41;06
Unknown
I would say 50% of the success stories that they were sharing was we built a business intelligence platform, uses SSR as a and power BI and Spotfire, and we showed 3 to 10% improvement in production or cost reduction or whatever the case. And, and there's so many companies that are still living in that world. Yeah. Where I'm like, it's great.
00;44;41;06 - 00;45;13;29
Unknown
It gave you like great results. But I'm sitting in my seat like on fire. I'm like, oh my God, imagine what you can do now with all the things that you have today. Yeah. And so when I started presenting and talking about the kind of latest and greatest technologies, what we're doing and all especially gen AI and, and compute and cloud enablement and adoption, etc., you can you can you can look at everyone's eyes in the room and they're like, we have decades and decades of, of, technical debt that, you know, and lack of improvements.
00;45;13;29 - 00;45;34;15
Unknown
And so those opportunities are amazing. And I can tell you big and small operators have it. Oh, yeah. I mean, but again, I walk in places and there's, you know, they don't even have a formal SQL server database that they're doing these things out of. And it's like even that they're running Production Application or SQL express. Yeah, I've seen it literally like access you weeks ago.
00;45;34;16 - 00;45;55;20
Unknown
Access right or without. And so and then so as a technologist you look at this and you're like scratching your head because you want to you want to make it brighter and shinier. Right. But then I think you have to be humble enough to say it's working for you. Yeah, yeah. A prove the point. It's solve the problem and the business is happy with it.
00;45;55;21 - 00;46;16;00
Unknown
Yeah. Imagine the cost of disruption, right. To move something from let's say, you know, there's still Hadoop clusters everywhere into, you know, a modern data mainframes still. Right. Solutely right. And so you kind of have to step back and say you know what. That's okay. Yeah I can help them. Maybe at some point in time or if they want the effort.
00;46;16;00 - 00;46;41;21
Unknown
But I come across these situations a lot and it's not it's not man, I have to stop myself from no say what I think we see in the Stone age. Like what I was asked special to. I mean, at the end of it, like, if you're getting the oil out the ground and you're not spending a ton of money, like, I mean, like, there's that's well, that's the thing that I think is the most interesting kind of promise with some of this stuff is that, you know, in the short term today, it's not going to take your job by any means.
00;46;41;21 - 00;47;02;12
Unknown
It's more than likely just going to automate a lot of the stuff that you hate doing. That is probably something about, you know, where I was going earlier. Some of you don't even think about that, right? Like there's there's still banging their head against, you know, the wall for days on end with an Excel spreadsheet or, you know, billion points or just have a record and they don't even think about using AI for anything.
00;47;02;12 - 00;47;22;25
Unknown
Yeah. You know, it's like, you know, I mean, that's that's the thing, right? Like the the future of these companies looks very, very different. I listen to thinking machines on our road trip to Oklahoma City, this past weekend. And one of the big things he kind of harps on is, you know, the subject matter expert is is the most important person in the absolute.
00;47;22;25 - 00;47;43;15
Unknown
Because at the end of the day, there is a long time period from whatever, 2 or 3 years ago when this kind of started to when all humans automatically are comfortable and accept and trust the output of these things. Yeah. And so up until that point, the subject matter expert is the one who has to make the decision at the end of the day.
00;47;43;15 - 00;48;12;20
Unknown
And they're also the one helping refine the model every time they give it a, you know, feedback. And think of it this way, like a lot of decisions made in our industry specifically is around reducing risk, risk of a environmental problem or a safety problem or an investment problem, almost everything. Right. And so I don't think we'll see models make those decisions for our business partners in anytime.
00;48;12;20 - 00;48;39;07
Unknown
So you're still going to need the same. Yep. You still need to like the human in the loop is so important and will continue to be so important in almost everything we do in the industries, given the regulations and the soil concerns. The question is, how much do you automate, right? And how do how much do you make those decisions visible to the subject matter expert on how you came to that conclusion, and then let them make the right choice?
00;48;39;07 - 00;49;03;14
Unknown
Yeah, that's how I tell everybody, whether it's from seismic interpretation or building a small little financial, you know, application that's going to. Yeah, look at, you know, whatever data. Yeah. Right. That's that's for me is is not going to go away anytime soon. Yeah. Well that's the thing. So like enabling that now allows you to scale the company in a much different way than historically.
00;49;03;14 - 00;49;21;21
Unknown
You would. Right. Like if you have someone like this regulatory automation stuff that we've done, they no longer have to have a person to do the regulatory filings, no matter how many wells they have. And that was a linear thing that before we added that automation, you get to a certain capacity and then you need another, you need another person.
00;49;21;24 - 00;49;50;00
Unknown
And so it's like the it's kind of crazy to think that you will have energy companies that are like tech unicorns in the future, where it's like a 50 to 100 person team of experts. So you can think of it this way, I, I've asked this to a couple of our customers, like, if you can start from scratch today, how would you build your company if I got first principles, kind of like yeah, like just think of it like if you want to build an AI first energy company today and you have a chance to build it from scratch, would you build it the same way you did?
00;49;50;00 - 00;50;18;22
Unknown
And most of the answer is no. Yeah, yeah, yeah. That's just that's the most exciting that sunk cost that I'll say in place. Right. You know, and I that kind of triggered in my mind when you were talking about, say some of these companies that have been laggards, is that actually almost a benefit now, like where it's at, you know, some of these companies have this technical debt and they have all this just in their mind, this investment and what they've been doing for the last ten, 15 years on the the they're trying to be on the bleeding edge of tech and data and all these things.
00;50;18;22 - 00;50;45;16
Unknown
And then again, some of them probably set really good foundations for themselves if they have good data behind it, and they can just layer on the AI and be off to the races. But like, and you can tell me if I'm wrong, I don't know how much you can speak, but like I know at GM we were pretty fairly tech forward for coming about like say we had a guy come over from Kraken and Crack, so wasn't doing much on that side, but now they're leaning in with you guys and like, I don't think they had a ton of like data or tech infrastructure, but now, like they can just lean right into the no.
00;50;45;16 - 00;51;04;09
Unknown
That's latest and greatest. You have kind of two camps there. You have what you mentioned where it's like, you know, a GM type thing where they've exited a prior company. Yeah. Now have funding and are now going to do the next one. And so they're literally saying, how do we make from scratch this company AI native. Yeah. Right.
00;51;04;09 - 00;51;33;02
Unknown
And a lot of the time that's it's not just because there's shiny new tools, but it's also because these people have been in this industry for long enough and they have been burned by these software companies that they no longer see. You know, the benefit from if you can do certain things with different AI tools because, you know, and then you also have the laggards who just never got on the the big data train, really, which they still have the data journey that they're going to have to go down, which that doesn't go away.
00;51;33;04 - 00;51;57;05
Unknown
But you're you are, in a sense kind of able to leapfrog in certain ways, you know, what you can do today if you because you're not married to another system, inertia you have to overcome or you don't have to overcome the change cost of moving from one to another or whatever, like that. So it's it is it's very interesting to see that it's exciting to, to see these people be like, hey, I would I do this from zero.
00;51;57;06 - 00;52;18;02
Unknown
Yeah. Right. Or like that's what's happening in all industries. Yeah. It's happening in all these. Some are faster. You know, I think like financial services have been phenomenal in adopting quickly technology that comes along, whether it's cloud or AI or data, you know, because I think there's not a lot of complex manufacturing processes that they have to deal with.
00;52;18;04 - 00;52;43;15
Unknown
And then you have to pay a significant amount of complexity and safety and environmental and regulations. And so I, I completely understand the logistics of it. But I think at the same time, the industry has been a phenomenal innovator when it comes to engineering. Yeah. Milestones and engineering, you know, achievements and, and drilling thousands and thousands of underground wells in the middle of nowhere.
00;52;43;15 - 00;53;04;18
Unknown
Right. And and with, you know, the it's just, I mean, the precision and accuracy of precision and accuracy and engineering, you know, chops are important. That's the most exciting thing for me. Yeah. And just my personal experience with this stuff is it's like, if you can empower a subject matter expert with the ability to go code and build tools for themselves.
00;53;04;21 - 00;53;24;18
Unknown
Yeah, like, holy shit, you know, but you're off to the races because most of the time the struggle comes from getting the SMEs to communicate all of the details and all of the nuance and all of the specifics and the edge cases to the project manager, who then has to turn around and communicate it back down to the business analyst like.
00;53;24;18 - 00;53;43;17
Unknown
And so it's just the communication piece, right? You're going direct to the source instead of going through a waterfall of different people that, you know, as you get further away, have less and less understanding of what it is you even do. Yeah. That that to me is like, but can you give now that reservoir engineer completion in an era, you know, they're a little sandbox.
00;53;43;17 - 00;54;00;21
Unknown
And it used to be, you know, a lot of people will just learn Python because they had to, but like, yeah, but but now can they, you know, give them the proper governance guardrails to work with them. But can they spin up what they need and then just say, hey, I need this deployed. Yeah. And I'm just build a staging environment then, you know, so forget about the wheel, right?
00;54;00;27 - 00;54;37;16
Unknown
Yeah. I'm I'm sorry. One of the things I think I told you this, you know, long time ago, what are the reasons why? You know, I left, Chevron at the time where I decided to pursue other opportunities was that we were working on a big, you know, data platform, Data Lake Initiative, for multiple years. It was around unconventional and, you know, primarily North America operations, etc., and he was taking us, you know, months and months and years of study, setting up the cluster and the Hadoop and trying to figure out how do you deploy and how do you store.
00;54;37;19 - 00;54;52;25
Unknown
And at the time, I started kind of learning, cloud and cloud technologies and services and then and it was like one of those lightbulb moments where I spun up an Azure data lake at the time and like five minutes.
00;54;52;27 - 00;55;14;23
Unknown
I was at nine. I'm looking I'm I'm like, okay, I understand there's no governance, there's no security, there's no nothing. But just the fact that now I have a data lake. Yeah, I can start I can start dumping data and analyzing it with a click of a button versus having to go through for months and months of, yeah, of rigor for no reason, really.
00;55;14;26 - 00;55;35;21
Unknown
Was was a turning point for me personally when I said, okay, this is and now I think the stage of GDI and the technology that's doing is another kind of huge, huge evolution of technology that I think is going to, cause a tremendous change in our industry and how we use it and how do we benefit from it and how do we use it properly.
00;55;35;21 - 00;55;57;23
Unknown
I even just think what I mean, right. And you said earlier, you're part of like migrations and stuff too. But like they just say about migrations like a, you know, say, if I was going to go from a Teradata to Athena or Snowflake to Athena Aria, but whatever direction you're going, like. Exactly. I mean, we could convert a whole code base of, you know, code from, you know, this dialect to this dialect or spark to this, or forget about like the all this modern stuff.
00;55;57;25 - 00;56;17;25
Unknown
I'm looking at Fortran code. Yeah. I got to figure out how to modernize it and actually refactored into rust or something like that. And it's like, okay, I was going to ask you that actually about, because you mentioned like tech that and stuff earlier, but like, because I, I mean, we're literally about to redeploy the front end that we've completely rewritten.
00;56;17;28 - 00;56;38;00
Unknown
Wow. React in like two weeks. That's insane. Because the quad code, like it is absolutely crazy that we're able to do that is my amazing, right? Yeah. It's amazing. It's just go away now because we have the ability to yeah do this I don't know. But it's one of those things you're going to it's going to accelerate some tech that right.
00;56;38;00 - 00;56;57;08
Unknown
It's gonna tech that it's gonna add a lot. But I think adding the tech that because so many people are going to be doing so much with the ability to remove it very quickly as well. I'm okay with that. Sure. Before people were experimenting, whether you like it or not, and people were creating technical debt. Right. And it was not easy to get.
00;56;57;08 - 00;57;16;04
Unknown
Right. No, no, it was horrible. Nobody wanted to do it. Nobody wanted even even like the simple thing of documenting focused on the documentation. I was thinking about that. Now it will document. Right. And so that's, that's the the paradigm shift that I think people in the industry, whether a technical or subject matter expert, need to understand it.
00;57;16;04 - 00;57;38;15
Unknown
This is now a different way of thinking about building solutions and applications and solving a problem. Yeah. Well even that right. Like for me, when I talk to our clients like the criteria I have for a good pilot candidate is it's repetitive. It's normally something that, is low to no value add to the company, but it has to be done.
00;57;38;23 - 00;57;59;00
Unknown
And the last one is usually the user hates doing it and it's like, yeah, documentation on code that used to be everyone's like, some people loved it. And these people hate it. Right. Because it's like you end up doing it after the fact most of the time. And then it's just it's there. But it's exactly it's a it's a necessary evil that you should do.
00;57;59;00 - 00;58;27;22
Unknown
Most people do very poorly because they just don't want to or it yeah, it. And so like those are perfect. That's a perfect example of like this is why it's useful to automate these things. Because now not only does a document generating a Readme file right. Like yeah, like here's the other thing that I tell my customers all the time is like a lot of these big challenges of how you you take it on and you, you make it part of your process and, and, more mature is that you don't have to figure out on, you know, right.
00;58;27;25 - 00;58;53;15
Unknown
Like, you have system integrators that will die to do this work for you. Yeah. You have cloud partners that they're only job people that are resources that are in these cloud partners. Right, is to do that for you. Yeah. So take like focus on the business value right. Forget about the plumbing. Yeah. The plumbing can be done. It's easy simple in a setting up the environments, the frameworks, the MLOps, the DevOps.
00;58;53;15 - 00;59;16;26
Unknown
All of this can be done right through your partners, whether it's the cloud or solution or the technology providers. Right. And, a lot of it sometimes is free. Yeah, right. So why not take advantage of it and then just focus on providing value to your customer internally? That's anyways, so what else? Yeah, we're rolling up on the WordPress on the hour.
00;59;16;29 - 00;59;31;25
Unknown
It always happens. It goes really quick. Yeah, yeah, yeah. This was fun to jump in on speed run real quick or yeah, I'll get creative on the speed round. That'll that'll work with I didn't know there was this fun questions. Well normally they're just like where do you like to, you know, what's your favorite restaurant here or whatever.
00;59;31;25 - 01;00;04;20
Unknown
Oh man, I can talk about this all day long. I want to ask what what is something you're most excited about and most not necessarily cautious, but worried about from, you know, a AI perspective in the industry. So that's a very good question. I am extremely excited about, the potential of vibe coding and, and, the potential of the tools that we have now at our disposal as technologists, as business experts, subject matter experts, etc., to solve problems.
01;00;04;23 - 01;00;29;01
Unknown
I think we are only scratching the surface. Right. I think of from research, you know, I was I was reading a document that was written by one of the main developers of OpenAI chat, GPT three, etc. the model. And it was just talking about the exponential growth and improvement in these models over the years, given the, the compute, acceleration.
01;00;29;01 - 01;00;52;26
Unknown
Right. And so and this was three years ago, and now I'm seeing it happening. Right. Like I'm seeing Asians take over the world basically doing tasks that nobody wants to do, doing a lot of automation. And I can do it myself. I have access to it. And my customers, etc. and I can see the exponential growth of productivity happening.
01;00;52;26 - 01;01;16;28
Unknown
Right. On the flip side, I think the energy demand of this technology is pretty significant. Yeah. And the numbers that nations are starting to talk about, you know, the hundreds and tens of gigawatts it has to come from somewhere. And if we don't do it properly as, as, you know, as humans, I think it will have a significant impact.
01;01;16;28 - 01;01;47;28
Unknown
So, I think that's what worries me a little bit. You can't not invest in it because these are matters. It's really a national security. So the security like, you know, this this will transform human, you know, life. And I think we have to be conscious of that enough to work on how do we solve it, you know, as, as, group of intellectuals and, and, and decision makers and politicians, etc. and, and I think that's an area of, of, concern for sure.
01;01;48;00 - 01;02;11;09
Unknown
Oh, so that's why I'm bullish. Natural gas in the short term. We got to get that from anything. All right I'm gonna tell you my options, right. Yes. I got more of a softball for you. Your favorite, Lebanese or Mediterranean restaurant and. Oh, my God. Houston. Oh, man. Are you Chinese? I'm Lebanese, born and raised.
01;02;11;12 - 01;02;34;03
Unknown
I moved here when I was 18. Two of them. And it's hard for me to say which one because they're different. One of them is a butcher shop called Musa. Okay, it's on Westheimer and Twin Health often. Fondren? Yeah, it's a butcher shop, but they have a little restaurant, so the meat is extremely fresh. Everything is amazing.
01;02;34;03 - 01;02;54;23
Unknown
There. They have one of the best chicken sandwiches you'll eat in your life. Nice. It's, charcoal grilled chicken. Okay. But you could also go and order their, kebab, mix. So they have everything. It's amazing. And it's just about the quality of the meat. Right? The other restaurant is Abdullah's. It's one of the oldest restaurants in Houston, I believe.
01;02;54;23 - 01;03;16;04
Unknown
Okay. They just reopened. They used to be on Holcroft and reopened a brand new restaurant on studio demand and I-10. Okay. How nice they do. They're one of the few restaurants who do, home cooked meals. Like, basically like house meals that, you see, you know, growing up in a restaurant, and they have phenomenal Lebanese breakfast.
01;03;16;04 - 01;03;42;14
Unknown
I think one of the only places that have good Lebanese breakfast. Nice. So those two, Abdullah's and Musa and it's Musa or Oh, USA. Okay. Yeah. It's an Armenian, Lebanese and so good. So good. If you like pastrami, if you like. So you can get some luck on the. Yes. Phenomenal. So nice. Nice. Yeah. I'm Armenian if it's, I'm half Armenian, half Lebanese.
01;03;42;14 - 01;03;56;22
Unknown
Actually, I'd say. Oh, it's good, but. Yeah, I mean, do you love it that know you have some guys from. What's my last question? What's something.
01;03;56;24 - 01;04;21;12
Unknown
Well, I don't want to pin that on you. What's your favorite vacation spot? Let's let's go with in the Middle East. Where would you go on vacation? I would always say my hometown. Yeah, regardless whether it's just Middle East or the world. And I've been lucky to travel so many places with work and personal. Bombing knows how much.
01;04;21;15 - 01;04;43;08
Unknown
You know, I have to travel to leave the family. At least in my early years. The reason is, it's just. It's just home. Yeah. What city is that? It's called Baalbek. Okay. Baalbek in Lebanon, and it's it's a mountainous area. Just just the simplicity of life, you know, you just wake up very different.
01;04;43;09 - 01;05;11;17
Unknown
Your family, your cousins, your people that you grew up with. You know, it has a pretty significant Roman temples and, and ruins that are huge that nobody knows about, you know, bigger than the ones in Rome. It was for some reason that nobody really knows a center of religious pilgrimage for the Phoenician, Greeks and Roman cultures for like 9000 years or 8000 years, whatever.
01;05;11;19 - 01;05;27;09
Unknown
And so it's a unique city, but it's super simple and. Yeah, and, you know, you know, everybody. So that's that would be my favorite. And then, of course, New York is another big favorite city of mine. And I yeah, it's hard to be in New York. You just sound the same that my wife in New York. I don't know, Rosie.
01;05;27;09 - 01;05;44;14
Unknown
I'm not there yet. Vance. Yeah. She's not from from Texas. She's from Florida. Right? Yeah. Originally, yeah. But we met in New York. Awesome. Well, no, I'm so glad we could, make this happen. Oh, man, this is awesome, I love it. Yeah. No, just, where can people find you? And, you know, can I connect with you on this?
01;05;44;14 - 01;06;12;19
Unknown
And I'm on LinkedIn. Hussein Shell with one URL is everybody? Yeah. Not not like the operator. And it was on purpose. Name it, but, Hussein Shell, I'm on LinkedIn. If anybody has a question or guidance or I do, mentor people sometimes when I have a chance. And so reach out to me for any customers listening to this, just please reach out if you have a problem we're trying to solve any partners, and then,
01;06;12;21 - 01;06;34;19
Unknown
I appreciate you having me, man. This is fun. No, I mean, it's just that whenever for a couple of years, you know, the pleasure. I'm glad we could do it. So, Absolutely. Just for the viewers, I think, like, know, Bobby and I have finally got some stuff scheduled, for some more podcasts, so hopefully we'll keep these rolling and, getting back on some, some kind of frequency for you guys, but preciate you kicking it.
01;06;34;19 - 01;06;54;04
Unknown
Of course. Kicking it off for 2020, 2026. Yeah. So awesome. Awesome year. So thank you I appreciate you guys. Yeah thank you. Thank you all. While some may see them as the crazy ones, we see genius. Because the people who are crazy enough to think they can change the world.
01;06;54;06 - 01;07;03;10
Unknown
Are the ones who do good by.