Or back. Second week of college football. I've got my slobbering hog on. We're undefeated. For the time being. You're not playing at Pine Bluff anymore, but Yeah, we we have our first uh test in Oklahoma City this week. I'm not super excited about it, but we'll see. Um today we've got Richard and Murat. Richard's the COO over at AI Driller and Murat's the CEO. Appreciate you guys joining us today, man. [clears throat] Appreciate you putting me first. First place, second place. It's got to be the hair. [laughter] I mean As someone who doesn't have any anymore, I do appreciate a a good set of locks. Yeah, I had to pay him 10 bucks to to announce me first. [laughter] Shh. Whatever. Get what you can get while you can get it. And appreciate you joining us, Murat, as well. You're in the Middle East right now, yeah? Yeah. Yeah, yeah. I hope my internet is good enough. Yeah, I think it'll work out. Sometimes I think it may look a little bit away, but I think the way that all buffers, it comes through on the with better quality and everything, so. So, our our new story for this week is uh I don't even know the names of the people, but shout out to them for fun ideas with AI. They deployed a bunch of AI agents in Minecraft uh and let them just autonomously start going to town and apparently they've set up all kinds of normal things like economies and bartering systems and trade and all this fun stuff. And so, given I mean, you guys have AI in the name, I thought that was a fun story to kind of lead into talking about what you guys are doing in in AI and you know, the the usefulness of AI moving forward, how people are going to start actually doing stuff. I mean, this is a fun, you know, concept, but it's also not the simplest thing in the world to do. So, it's interesting to see kind of how people are approaching these things and how you can translate those into real world use cases. Yeah, so for us like honestly like it's fun but there's a fun story about the name on the AI driller in general. Um but um you know, overall like what we see with AI especially in the industry things get overused, right? Like we were talking through 2016 big data was a thing. Now then AI became a thing. Now GPT is a thing, you know, but uh um so this was like sometimes you know, like it actually burns us in the name. People are like, "Where's AI? What's What's AI? What's AI?" Everybody expects something uh something out of it. But uh coming back to the name perspective, we were working on the project with Shell at the time. And um we need to come up with a name for the project. Uh we were working on automation of control system for the rig. I'm staying on the floor with the driller uh using my phone trying to find a free domain name for the project. And uh [laughter] on the AI driller on it. And I was like, "Man, that sounds pretty fun." So I talked to driller and said, "What do you think on the name AI driller?" He's like he's a driller, right? He's like, "Man, this is the best name I ever heard in my life." What AI means? You know what I mean? This is exactly [laughter] And uh so this what became the name of the project then became name of the company later on and became name of the product on that front. But that's actually People always think that names from companies come from some kind of sophisticated thought process or it's always like some craziness and you're on you're on a Google or you're GoDaddy or something and you're trying to look up all these names and it it takes it took me like whenever I started my my first business, it took weeks for me to come up with my my official name. So kind of funny hearing that story. He asked the driller if he liked the word AI driller. Duh. [laughter] No, I think uh that's a perfect it's spot-on because it's like the people who took a lot of time I in my experience at least you end up running into like oh well someone else has that cool domain name too. So now I've got to go back to the drawing board and find something else or find a weird you know dot AI or IO or whatever. [snorts] I wouldn't be surprised I wouldn't be surprised if there is like a bunch of AI bots running through all the free domains and just buying them up by bots you know what I mean. Yeah. To resell later. Some of the names you can you know they hundreds of thousands of dollars right now. So many so many like with with that like was was or is artificial artificial intelligence like a core part of what your business model was or did it just happen that I mean cool name and you know you maybe knew ultimately you want to kind of involve some elements of that. You just kind of curious cuz I mean obviously since 2022 December like AI has taken a whole new different meaning for people since before then. So just kind of curious where the genesis was. Yeah I mean honestly like you starting a company right so the first thing you're going to meet with a bunch of investors and and everything else. So AI becomes part of your decks like what are you guys doing differently? I'm applying artificial intelligence applying certain ways of doing things. And of course like we work in automation world so you have to you know consume a lot of data real time and make decisions based on it. So there is a you know certain levels of mathematical like processing will be involved like you know. And AI is very broad. What is it what does it mean artificial intelligence you know based on what right? So there is multiple machine learning techniques being used to process the data clean it up structure it in a way to you know to get insights out of it. You know and then and then this is our part of our business day-to-day. At the same time we do a lot of things which are not AI like in a drilling world you need to design a well. And you know AI has tend to you know like tend to change the answer. In engineering world the answer have to be the same every single time with the same inputs. You know, so in the in we rely a lot on physical based models as well as you know, normal mathematical modeling. And you know, when it comes to you know, the answers to your CEO or the operator of you know, E&P company. He needs to know like what exactly my budget was, what exactly my actual cost was, what exactly like how many wells I drilled. AI can be used but he needs exact number. So then the day if you're going to be audited against those numbers later on, you having so ideas like there is certain aspects of things when you know, the KPIs has to be based on actuals. You know I mean, like the engineering outputs big have to be based on specific design outputs and you know, AI isn't used to you know, make data more accurate and you know, in across all those aspects of things. And um this is why like I think one of the biggest things from what we were excited with you guys on the Collide GPT specifically, you know I mean, it's just it's not the overall good good boy answer. It's you know, specific tailored to the industry uh platform when you can actually trust with much more you know, I mean, like uh confidence because I'm still a drilling engineer. Because you know, okay, it's not just a bunch of different stuff which you know, drilling drilling could be drilling a wall to frame the picture, you know I mean, like uh it actually understand what the drilling and from the SPE papers and everything else from there. So that's gives you much more confidence. Okay, it's you know, like it's learning from what I would learn as an engineer. And this is kind of like I think in our industry AI is part of everything nowadays like part of Excel sheet, part of you know, where we write emails. But it's the answer is always like what's what's the benefit? You know I mean, always what's the value I'm getting? Okay. 100%. Um So then one thing I got cuz I mean I'm I'm at an operator, you know, currently and just kind of curious, have you actually had headwinds because you have AI in your name, especially since like GPT cuz I feel like there's clearly a lot of people that have been trying to capitalize on AI. So I think as soon as some people hear AI in the name or just even in the pitch deck, they're like, "Oh, these guys are just trying to capitalize on it." I mean, I'm just curious like how do you How's it been trying to break through even just to get the meetings and then um and then we can maybe funnel that into like getting buy-in from the consumers, you know, on the product that like all this automation's not going to steal your job, it's actually going to make you better at your job. Um and just trying to get people's head wrapped around like being more comfortable with using that kind of technology, um especially in high-risk environments like drilling completions. Yeah. Yeah, and it's like uh we have black boxes, right? Yeah. So that's that's what AI has been in the past. So how do you overcome that? Cuz you guys have gained a ton of traction. [clears throat] And that's that's one of the first questions you always get is, "Where's your AI at?" [laughter] So yeah, you have to cloud. You have to just talk about where your value's at, you know, talk about the automation that is in the system, you know, the rig state detection, some of the some of the machine learning and and the automated math and the one-click buttons to get stuff done. So um you just provide value, you know, when people first ask that, I always I always want to say is, "Where's your AI at?" That's always what I want to tell people. Everybody always talks about AI like they have AI all around their life, you know what I mean? And sometimes I don't even know what AI is and I work for AI Drilling, you know. Yeah. So So like Well, yeah, I mean, I think you even there like the lines have been blurred so much between AI and machine learning, deep learning, but even just going back to like um just more direct methods and it's like even just if you're baking in to your point, Marat, like SPE papers like, you know, you know, different, you know, uh academic papers and concepts and you're actually just providing those. I was like, is that AI? I mean, like I mean it's it's definitely it's taking someone else's knowledge and applying it, you know, programmatically, but I think it's probably up for interpretation, but just, you know, maybe even at that highest level like you know, where do you all draw the line between AI, machine learning, you know, and uh uh So, let's throw automation throw that word in there, too. Automation, I hear that one. Automation, yeah. A lot. Yeah, like I mean kind of the original question would be like, uh did it help us? Um it's kind of like yes and no, right? So, from one side like it was a is a word overused? 100% Overused for all reason? 100%. Is our industry educated enough what it actually means? Not at all. You know, I mean uh from that side. Nobody knows what it means. Nobody actually studied any like, you know, like there's very few people who had like went down and actually did a data science not 101 course. Like I'm not talking about writing a some code and doing any sort of, you know, initial training from that. So, all of industry uh companies went through the whole data science department phase when they hired data science department and keep them on the floor and say, "You do some AI for us." On their fronts of the we went through those phases, too. So, I did like uh word definitely overused like I just said. We come up with this AI what AI means. We had twist of like AI is automation intelligence. AI is all inclusive. AI is this and that. But at the end of the day like it's kind of funny that out of AI jeweler you know, the way our customers call us, it's all AI. So, like even though that's the same two letters which were kind of like, you know, you said like been overused, this is what the name of the company today like, you know, if uh they refer to us and hey, call the AI guys. Hey, uh did you see the AI guys? You know, I mean hey, have you tried the AI you know, like AI platform? So, like you know, the way you word which we we worried about, it's just like the name of the company and name of anything we do from that side, you know, like we shorten the name everywhere on mobile apps and everything else from there. Sure. But yeah, I mean it's branding. I mean you And then again, I I I think to Richard's point, like once you just demonstrate you can provide value, they don't even care, you know? And again, it's it's a memorable name, so there's that, too. I And I I I don't want to focus too much on it, but I just know like you know, again, coming being someone who gets sold to a lot or I have people even just jump into my, you know, LinkedIn messages all the time about this and that and the other. I mean, you're just getting barraged with, you know, AI this and imagine this and then you have all these things. So, just I'm just curious like I'm sure you guys run into those headwinds and uh Well, we always try to What I always try to tell people is like we're getting we're trying to make it better better, right? Like automation, artificial intelligence, AI, whatever you want to call it, it's uh takes a long time for it to get to where it needs to be, right? People want it to be uh here yesterday and I'm like, well, look, we still got cars that uh aren't fully automated and they've spent probably trillions of dollars if not hundreds of billions of dollars trying to do this process. You know what I mean? With big-time investors, we don't have that in our industry. Software industry within oil and gas doesn't have the the money and the finance, you know, Marat had to bootstrap all this stuff by himself, you know what I mean? So, uh you know, it's not going to be an overnight thing, but we do automate right now. We do automate people's workflows uh throughout the entire operation and we will make it better and better and better to where people are legitimately pressing one button or talking into something telling it to do to do something, but but people got to understand you have to be a part of it so that it can get to where you want it to be, right? So, Yeah. that's that's what we're doing, so. Nice. So, I guess probably from my own personal notification and then just obviously for the users, I mean, the way I kind of want to slice this and y'all can tell me if you want to go a different direction, but like let's start from say like the front end that you have now, like how people interact with AI Driller, the value it adds there, and then maybe we can work back down the stack, you you or the process and just talk about maybe some of the data cleansing and stuff that happens and even the data ingestion, you know, upstream of that. Um, but if you want to like jump in like, you know, what does your platform provide today and like how do people interact with it on a day-to-day basis? And who are the consumers? I guess, you know, is it a directional driller? Is it a drilling engineer? Is it all of them? You know, all those kind of things. Yeah, so main idea for AI driller is that we are communication platform across multiple, you know, like uh multiple stakeholders. So, today AI driller used to if you go to like traditional drilling site right now, right? Uh you have a, you know, like uh multiple vendors using completely different software packages sitting, you know, working the same goal but sitting different uh packages from that side. And uh uh with introduction of digital world, you know, our demand from field guys been increasing every day. I want to see this metric. I want to see this KPI. I want to see that KPI and everything else. So, it's just like from your field perspective, your amount of paperwork day increases, increases, increases more and more. Then you pile on top of it what your customer requests from you to be tracked. So, like it's uh And how do you go forget uh the the speed of drilling is increasing, increasing, increasing, too, right? So, it's another you're stacking stacking stuff on top of stuff, you know? Yeah, so for sure. And and like just so so I think you're here, too. I mean, like is this onshore and offshore, you know, international and domestic, you know, lower 48? I'm obviously Marats in Saudi right now, so I'm assuming they all are trying to Yeah, but there is a misconception later. Drilling is different from onshore to offshore and uh you know, longer well, conventional well, wells and conventional wells. At the end of the day, all we're doing is making a hole in the ground. You know, whether we do it from the platform, from the little video rig from that side. Um, but what we just kind of mentioning that efficiency what we see in unconventional, like we can't afford any like lack of performance in there. And uh that increase of efficiency has not been uh updated on the workflow perspective. Like, you know, I do this like if I'm a drilling engineer, I speed out today you know, I mean, like 15 to 30 procs uh if I only run one rig. Back in the days, it was three, four programs I need to create a year. I'm like planning engineer right now, then I'm a operations engineer, then I go into review stage on my end, but uh nowadays then each customer like, you know, what they goal is like how do we do I scale up? How do I do more with less? You know, I mean, how can I my engineer do three rigs now? Meaning it's 90 procs in my year. Think about like 30x improve in a change of actual reporting capabilities from that side. And then the that kind of brought one thing, and the same thing came with a, you know, like if if you are drill site manager on the rig of what's happening with you. Like, now you have to plan for next well coming up there. Uh Midland is where where where our office is there right now. Like, it's probably the most efficient place to drill in the world. Like, you know, especially being in the Middle East thing in different aspects of like logistics and everything else. Like, we don't appreciate what can, you know, like things like Texas provide when they just do a call and everything is available for you almost same day. But uh uh what's happening from a lot of guys perspective, they rigs nowadays and it's just all the same thing over and over and over and over and over and again. And uh and we what we've done as industry accumulated a lot of data. It can be now used to completely automate the workflows with the guys. And this is what they are doing and kind of come in. So, first of all, we bring everybody on the same page. We remove completely the paperwork from every single key stakeholder on the rig side. Uh everything falls in. And ideally, like sometimes it workflow is we pull from sensors and create a reporting for them based on sensor data. Sometimes the AI is used just like be able to just go back to 400 wells long while like let's say you are in big company and you bought a new asset and it's completely like you know a bunch of folders and thousands of wells you have no idea what to do with it but you have to put a budget for next year what I'm going to be drilling. So this one AI will be used to process all that data and parse all those thousands and thousands of PDFs and you guys when you training [clears throat] collide you appreciate like differences between reporting of one vendor versus another and everything else from that side. So it's anywhere from like data processing and data uh data processing to like image recognition from that side. And also like then we Yeah, I think that's one of the big things that I want to point out is like we've been talking about AI and stuff and yes you guys have enablement via AI but there's so many things especially in the oil field especially on the services side where it's like just combining workflows or centralizing data or stringing workflows together like they are in in the real world but in one place brings in inherent efficiency and optimization right out of the box right which is why I think there's been so many of these cloud tools that have popped up in oil and gas and you know whether it's on the production reservoir drilling or completion side that gain so much market share because it's it's not necessarily that they have like a ton of secret sauce and I'm not saying that you guys don't have any secret sauce but just by putting it in the cloud from the traditional platform or workflow that it was you're enabling these integrations these you know stringing of workflows together these automations and things like that that in again inherently optimize and speed up a lot of this stuff. And then you layer ML and AI on top of it to make it even more efficient, faster, automated, and and things like that, right? Yeah, and uh Bobby also to to answer a little bit of your question, pretty much everybody on the on the rig site is going to be an AI driller. From MWDs and all those surveys, the to the directional driller, all their BHAs, pre and post. Um to the we have a full directional drilling suite on there with anti-collision, all that kind of stuff, geosteering. [clears throat] Uh the company man is going to have like pace on view on there, parameter comparisons, road maps, all that kind of stuff of performance. Same with the drill and engineer. We have well progs on there. We have um full-blown engineering modeling suites in there for torque and drag and hydraulics. Nice. Uh and then for management and all that stuff, we got all kinds of like KPIs and pretty much anything that you can imagine that you'd want to to look on your organization, say, "Hey, where can we get better?" Uh we have, you know, KPIs that we can build for you guys and help automate the workflow for for management for, you know, like their quarterly reports, monthly reports, their daily reports, stuff like that. So, pretty much everybody within the I I when we say operations platform, like it is an operations platform for the entire operation, so. That's pretty wild. Um as far as So, yeah, again, let's if we're going to focus on the front end and work our way backwards, like um so, like you said an app, so you guys have like iOS, Android apps on top, and then also like, you know, is it browser-based web apps or, you know, how does that that all kind of play together? Yeah, from from one end perspective, like kind of continuing on John's point, like all like cloud infrastructure and you throw in apps on every platform, you know, I mean, like your customer working on like iOS, Android, you know, web apps and stuff like that. Uh we do do uh couple customizations here in the Middle East as well just to, you know, make them more available through different infrastructure and requirements, you know what I mean, because we have when you ask when everything's public cloud and everything else like that, not all part of the world support that. Sometimes you have to be on private environment and you know, just depends what possible. Uh in the end of the day like we solving one issue like you know, John mentioned this communication issue, right? By bringing everybody to cloud, people can communicate freely. When something change, everybody knows, everybody on the same page, nobody say I didn't know, I didn't get the email, something bugged out. I did this, you know, so it's kind of like make sure that we all become part of that digital economy nowadays. Like you know, because in the end of the day like it's just you have so many technologies surrounding your day-to-day life. When it comes to the drilling or completing the wells, it's you know, a lot of those technologies and just not around you. And so you have 2 weeks on with technology, 2 weeks off of technology when you go to actually on your 2 weeks on. You know, gas from that perspective, but we still work with a lot of machinery from a lot of different things from that side and uh More tools just easy for the guys like I feel like we always feel like when you know, technology going to come, but what we as industry start realizing is uh the technology is now day-to-day life so much more advanced. Like we already all trained. We don't need to be trained on like how to use mobile, how to use this, how to log in, how to you know what I mean like make sense of KPIs because we you know, like the watch now track our KPIs so when we walk, it recognizes when we walk, when we sit, when we run. Calculates all of our like you know, like different sensors like the little watch on your wrist right now does more than you know, most of the systems for drilling. So like and then then you go back to work and you know, you start typing your report again. You know what I mean? This is kind of like your uh main uh I think one of the main kind of approaches to from the industry like people already does not mean uh meaning mean the age. Like you should see like you know, the 60-year-old company man. When you know, like you would bring the automated report, they're like, "I'm so glad to see it before I retire because I've been wanting it for 20 years." You know what I mean? Like, so this is just like those kind of things solve the most issues and it doesn't matter how they even done. You know what I mean? Like, even uh it's just is it the eyes of the AI GPT or whatever? As soon as it solves my issue and I can rely on it, I'm happy. Like, I don't care what they, you know what I mean? What actual algorithm been used, how it would have been done, and everything else. It's like at the end of the day, it's all about does it make my life easier? You know what I mean? Like, that that is it. Free my time. You know what I mean? Is it convenient to use? Is it reliable? Is it scalable? Those are the questions. The answer is not about is it run on Kudos like Kudos Kudos video card or it's like, you know, what what's this behind? Is it H cloud? Like, why should I care? You know what I mean? Like, for for my side, this is my main approach, okay? Yeah. So, [clears throat] obviously to to to translate for you, Bobby, it's this cloud-based. And we we do have an offline version, a desktop version called AI Spaces that Correct me if I'm wrong, Murat. That's a what? C++ type of thing? That's like our backup just in case anything ever happens. It's syncable and all that for for the EDR data and all that kind of stuff. Um uh C++ on that, Murat, for Spaces? For well planning and directional drilling software. Okay. Yeah, so that's the another software we don't really talk much about. But we do have like a full-blown directional drilling software. Okay. That's pretty neat. Yeah, I mean, cuz I guess cuz like I guess some of the some of the things that we want to dig in here cuz I mean I I agree 100% like on like the business value and at the end of the day like company man or someone sitting at a desk could couldn't give two shits if you're using React or you know, plain vanilla JavaScript. But at the same time, I mean, one thing for John and me for our audience is we want to we want to get into like you know, like what it's what tools are being used that enable you to deliver a product that the users don't realize is, you know, that they're using it cuz it's just such good, you know, software. Um so, I mean, I think with that like to your point of like, you know, you do things in environments like say in the Middle East where like there's data sovereignty and everything like you can't say, "Well, I'm using AWS US East, you know, one, you know, so so deal with it." And they're like, "Well, we don't want to give you money, so deal with it." So, I mean, maybe how have you architected, you know, even just your dev process to be able to, you I mean, is are these containerized? And then, you know, say you know, the fact that you were able to deploy apps say on Apple and the you know, web browser and all that like are you using say something like React Native where you have to code once and deploy anywhere or is this more like do you actually have like native iOS developers um and so on? Yeah, from from that perspective is like uh when we talk about stacks, right? So, stack is um not as important not important early stages. Um yeah, but um the most of decisions you make early stage your architecture going to live with you later on. It's just like in the fast pace of the startup, you don't won't have time to go back and refactor everything and change like So, sometimes you just married to certain stacks as you go in like as you grow as a company and learning new things, some of the things will be there behind you. But uh one of the things what we do on our end is uh when you start changing quite a bit like you have to nail that data model and approaches early correctly um because later on that what we will be your limitation of the system. You know I mean, for us we were yeah, a little bit later entrant to the um to the industry when it come to an analytic side and operation side and uh uh this was a kind of biggest advantage. So, we knew what's missing. We know based on feedback from the customers and everything from that side. So, the early decisions we made still, you know, makes us, you know, like a kind of one of the front front end systems going going further than anybody else at that time, but a lot of this stack will start changing when you start streaming 200 rigs. You know what I mean? Like yeah, like yeah, the optimization become an issue. The first level of optimization, you know what to do, and sometimes you will feel like you already went through everything possible because at the end of the day nobody cares what stack you use. It has to be absolutely fast. You know, like everybody compliment us on fast. I usually hate on sales because I feel like it has to be faster and then faster and faster from that side. So, then we can start changing a lot of the stacks trying to figure out like what uh segregation of data can use for time series data, what can I use for my logs, what I can use for my metadata. [clears throat] So, system becomes a little bit more agile from that side and uh [clears throat] From our side you know, like we use Java on the back end, you know what I mean, from from that side just pretty much enterprise since we're working with enterprise customers. And uh I apologize, guys. But uh uh from front end side like you know uh normal JavaScript and you know, pretty much cloud base is a lot of libraries you keep using on that front. And initially those libraries get you started quicker, but uh customers like to have customization. And that customization walk you away from libraries very quickly as well. So, initially you start with something you can just configure, then you throw everything you coded for 6 months away and you build from scratch, you know, everything else, but you cannot start there because you never know what customer want in the end. So, you want to move very very quickly so they continue in in a continuous manner, get the feedback. And once you feel like no more feedback come around, then you just go back and redo from scratch from that side. So, this why architecture from our is just more like go as fast as you can to make sure customer you don't lose the customer's idea. And then once you, you know, like do it, then you go back and start cleaning it up a little bit in a minute from that side. So, that's very important time because if you don't clean up at some point you you know, at the bigger scale you start failing and then from Yeah, that tech debt doesn't go away, right? There's some baggage that comes with you. Yeah. And then what would be the I've got a follow-up question on that, Murat. What would you what would you recommend or what's some advice you would give to someone who might be out there that, you know, has an idea for a software startup in oil and gas? What advice would you give them specifically around you know, architecting with the data sovereignty piece, right? Because a lot of people stateside have no idea that, you know, basically outside the US it's a completely different structure. There's different regulations, there's all kinds of different expectations and data standards that come with that. Like the data physically cannot leave the country. So, it has to be on servers within that country. So, your plan of you know, using AWS or Azure on your, you know, EC2 instance in the US doesn't work anymore. And it also prevents you potentially from securing very large big contracts with the big IOCs which look great from an investor perspective and as a startup you want to get, but then it it comes with all these stipulations, right? So, like what are some lessons learned or advice that you would give just based off your experience? Based on my experience will be like if you just getting started, don't worry about it. Because in 5 7 years by the time your product is solid enough because if your product not solid you have nothing to do in Middle East or any other countries because you're going to be failing very quickly and it's going to be expensive from the get-go. You have to get certain stage of the company to even, you know, start thinking in our geographical expansion. The beauty of US and economy like when the users can make decisions very quickly and you know, it's transactional economy. Uh this is perfect time to grow. By the time you grow in 5 7 years, my personal belief, everybody will switch to public cloud. And I deal like it might be certain like, you know, limitations, but those limitations will lay on the shoulders of Google Cloud, Amazon, Azure. Those guys will be carrying those safeguards and everything else. But this way I feel like from Ness standpoint, just align with a company like and go from there. From other side, what we've done to be able to do it like at this stage when, you know, not every country switched yet. Um we utilized the absolute minimum of the available pre-built functions for us. Like, you know, on on the cloud services. So, we kind of used our own stuff. So, in this case, we stayed independent. Like, we universal. We can deploy tomorrow on Google if we want to. You know what I mean? Like, if we you know, it doesn't change our code structure or anything else like that. It's just, you know, like it will be mentioned. Just, you know, you so containerized. We don't care what the name of the server is. And you know, like sometimes you talk to customers they're like, they don't understand in details. They're like, "Oh, we are AWS." Oh, cool. You know, I'm in AWS, too. Like, you know, "Oh, I'm in Azure." Like, at the end of the day, it's just freaking server. You know what I mean? It's almost like "What TV do you use at home?" I use Sony. "Oh, I'm also Sony TV." Like, this is like this is a discussion we having. Like, you know, it's I know it's sounds like AWS S3. [laughter] Like, it's it's just a storage. Like, it's SSD inside inside there somebody's data center. That's all it's called. Like, "Oh, I like S3 buckets." You know, it's You know what I mean? It's just means absolutely nothing. Like, you know, it's At the end of the day, we have the server which has to have a memory, processors, you know what I mean? on that front and and uh kind of like a database and it's better like work perfectly and have like a good you know like a reliability standpoint but This is why I feel like sometimes we just over complicate also add that you probably want to be financially prepared for for something like that cuz it's a long process be patient and um you know it takes a while not nothing over overseas is quick fast so you have to have money in storage to to be able to come in there and and hold off for a while to earn earn those contracts so No for sure that's a good a very good point that I feel like a lot of people don't or at least outsiders to the industries that come in you know their top targets are always the super majors and the IOCs right because they see okay well they've got the budget and it's like what you you're going to spend two years trying to get into one super major and you're probably going to do a bunch of free [ __ ] and they're going to hammer you on price and then maybe you get the opportunity to participate in a bidding process and then yeah but you you're like spot on there's so many companies that come in and they focus all their resources on the big budgets well they run out of their own resources before they can secure any of it and then they're shut down So I say I would say you know build a good solid product that people like and then then at that point in time you can start to consider it you know not only do you have to build a product but then you also have to stack a little bit of of money have a little you know excess to to be able to do that type stuff so It's kind of funny like when you doing that process build up on your point John is a So if we're going to work uh Sorry go ahead Sorry guys I'm a little bit delayed so if I interrupt anybody I just feel sorry but ideas like uh kind of building on John's point we when about big budget right There's what misconception on startup oh, this company is trillion dollar company. They can definitely afford $1 million on my software. Like the budget means absolutely nothing like what the company size is and everything else. We have to understand like it's almost like a person. The richer the person is, the more he counts the money. You know I mean like idea like some of the most of the companies doesn't not mean matter the size. What the conception was of them like being 3 4 years ago, 5 years ago like every company right now focus on cost. You know I mean where this industry need to figure out how to do it cheaper, more efficiently and stuff like that. We've been, you know, successfully doing it for last 5 6 years. This why you have to always strike and remind that so cost of software will be dropping down. You know I mean like uh uh um because idea like I don't know if we part of that same focus of customer You know like uh I want to reduce cost overall and we are cost for them, too. Of course we saving them a lot of cost, but at the end of the day we are still charging for the for the subscription, right? So from the from our end like uh we always think like our value as a company goes down. You know like probably 50% every year what you know I mean like from that side. This why you have to build twice of a product by year end just even maintain your pricing. And uh this is mindset of for the guys in that side like you know first of all you have to uh love your uh idea again for very much because it's going to be a lot of things like especially oil and gas. It's uh hard market to penetrate initially. But uh the demand of customer is coming from home, not coming from the industry. It's like I want to be able to be see it on my phone. I want to be able again this is kind of minimum requirement. You have to have a mobile apps. You have to you know like you people bringing their home experiences to the demand on the softwares today. So before you even you know consider to be valuable like any survival product, you have to do all those things together even in the US. And then you have to do three times those in you know any other country. Well, you make a good point, too, right? Is that it's you're always essentially behind, right? Like even if you have a great game-changing product today, in 6 months if you don't keep innovating on it, you're behind. Cuz someone else is going to copy you and keep pushing or they have bigger budgets or you know whatever it may be. And so that's a No, I empathize with that a lot as we keep making progress on Collide. What do you think of this competition? This competition is always like I know I know absolutely nothing. All I know is like if we go to sleep, some of them might be awake. So, like this is why we don't go to sleep, you know what I mean? And this kind of Well, mentality should be in the company like you know at the end of the day, we're serving customers, not competition or like what market shares, what different markets look like. Uh our like segmentation of the market in the industry is so small for software, like even thinking like where do I play, where I don't play, is just you know like useless exercise in that perspective. All we need to think of like how much time do I save them? Uh what's my actual value do I bring? How much excitement do I generate, you know what I mean? Like and that excitement going to go away, so you have to maintain that excitement throughout. And uh what's next we will be doing, you know what I mean? Based on what we're learning every day, you know? So, this is kind of best way to go. I mean I mean I always tend to say um when people are worried about sharing their idea or you know, they're really stealthy early on with the company. And I think there's there's probably a time and place, but at the same time it's like if you give someone your idea that you've already executed on it and then they do it better and faster than you, you probably weren't going to make it anyways. Um it's kind of my two cents. I mean like your point, you just have to foot on the pedal and keep adding more and more value, or else you just you're not going to make it. Until you're get bought by Landmark and then they can just keep selling berries and open wells for Sorry. Haliburton, I'm sorry if if you ever you know, [laughter] I've been a little rough on you. It's been a rough week. No, I mean, you have to keep in mind like from that side is you know, a lot of these companies will start developing internal tools as well and everything else because and it's a good exercise. I feel like every company has to have some sort of development going just to understand, you know, like the nuances and differences what it means and it's just good exercise when everybody learns together. Uh but what industry showed to like in the last, you know, like uh probably 10 years. My first encounter like my first automation system I work at drilling was in ExxonMobil. It was 2008-2009 time frame like 15 years ago. Uh we were drilling on Sakhalin oil wells like, you know, longest wells in the world at the time. And the idea behind was like it was it had an autonomous system for mailing, so I can send my email as soon as my subject line is like perfectly aligned, there's a mail server which going to process it and push it to the right folders. And then sometimes do it like more more analytical parts on it like 15 years ago, you know what I mean? So, a lot of companies were working on something, especially bigger companies I think. But now all of those companies could learn the difficulty, they learn the competitiveness of this. You know, you can get the product within a big company to certain level. You get excitement of all your guys in, but then the development team move to another more critical project and product dies. And as gets frustration the team always spend, you know what I mean, like a huge budget on the development. I don't want to, you know, get anybody in. So, your team has to, you know, like get with a very average product and end up doing this. But the majority of customers nowadays that we we see switched from that internal development mindset of like, okay, I want to have more market type competitive products and because I can also learn of on the features requested by other customers, not just from my own team. And uh this has been a big transition I could see it in the internationally, the same is in the US. Um they will pick up um more and more analytical companies in and just try to integrate to the workflow, see what sticks. And then what sticks become the part of the digital strategy and transformation. So I I I didn't want to get in cuz you you mentioned it, you know, probably about 5 10 minutes ago, but about like data modeling. And I if we were going to work back down say like the stack and just kind of think about that. I mean you guys it sounds like you're obviously getting data from all different sides of the operations, right? You know, you were just talking about MWD and then even all the way through to, you know, PDFs and historical data and everything and so just kind of curious how how your data pipeline looks. I mean like you have to have a staging layer and is it more unstructured with a MongoDB or data lake kind of thing and then you have an application database that's uh more rigid or more, you know, for lack of a better well-thought-through like data model that's, you know, probably more normalized and everything. Just kind of curious um how you guys, you know, handle processing that data um you know, from all the different sources you get data from is a workflow that is almost like different workflow on our end. Depends what they are what it does, you know what I mean from that side. Obviously files like any like uh object type structure will be, you know, [ __ ] on our end like, you know, this is how we do logs and audits and everything else in the AI. From that side when it comes to like just big files like, you know, we use, you know, a lot of uh just raw normal buckets like S3s or anything just at least to store it up so we can, you know, refer to that file later and then you go through parsing exercise when you structure it into your data model from that side. On the time series we had to build from scratch our our own way to handle data and this we call it BI cube technology that also allows us to pull, you know, thousands of wells of time series data of any piece of it, you know, within like, you know, a few seconds on that front and you know, we definitely cannot rely on any sort of database for that. So, like uh we had to do internal uh full development on how we handle, you know, 1-second data on that front. And uh comes to like MWD data, some of that is sub-second even. You know what I mean? Like so, we had to go back and readjust our assumptions on the time series to go sub-second on that front. And uh so, it just depends on the on the on the on the back side. But as you as I said, like on the small scale, you start with anything and you start changing when it's uh gets slow, you know what I mean? And then the that's the best outcome because uh more technologies you're going to pick up front, uh what's going to happen like more likely 2 years down the road, you have more ideas from other companies or anything else like that what's going on and you go to conference and you see some guys dealing with it much more, you know, like elegant way. So, like uh that's why I just start with what works right now, adjust as you go, you know what I mean, in the front and uh just kind of go from there. That's Yeah, I know what you mean. No, I mean that makes sense. I mean, it's sound like you guys are just very practical about about how you how you handle that. Um [clears throat and cough] And then as far as like say ingestion and you talk about the real-time ingestion, I mean, and again, like are you guys you're what what percentage probably 90 plus percent is on the drilling side and then you're doing some Are you moving into the completion space or are you doing a good bit on the completion space cuz I wanted to maybe get in about I mean, I feel like drilling has been ripe for these types of innovations because they have had that standard of what's a mall where uh the completion side has much less of a standard. I know they've been trying to push uh completions with some all standard, but uh I'm just curious on your thoughts like has that been like something that's really enabled for what you guys are doing? Um and are you seeing some of those headwinds on the completion side? Yeah, we entered the completion completion market last year. So, we you know, like uh build a uh completion part where I drill 53 on that front has been growing as well on our end. From us is like a, you know, drilling is not because of its XML we doing all this stuff. Uh, with XML structure it definitely helps because everybody bought into this idea 15 years ago and it's working. Uh, a lot of the things we do in despite with XML because it's not like it is a most, you know, like, uh, best way to send data back and forth and everything else that [laughter] is a multiple efforts to go to like different better XMLs that have more objects and there is still question is it the the right way to do it and stuff like that. So, I think uh, we, you know, the drilling industry is just kind of easier because of, you know, with XML and also despite the XML. And just if if it if you know what I mean from that side, right? Uh, from what it helps with all portion, what it helps like company like us, if you're a software company, you just all you worry about your customers and, uh, writing code, you can be heavy like us, like you're 80% on the development side, 20% customer facing just to like, you know, like, understand the features better, train people, onboard people, but like you go very, very fast on development for portion from that side. Uh, and with the completions right now like because there is no standard, uh, there is no even like standard way to send data from the field. You have to become from software company, you have to become a logistics company. Because, you know, you put your bunch of edge devices and you drive to the rig or to the site, connect, Become a babysitter. and then you call them and just to double check that what channel one is, what channel two is, what channel three is. And then, just because of all this logistics, you cannot charge a software cost to the customer. You charge them extra consulting cost because like you have to have a person available somewhere around if something goes wrong. So, I I don't think like a completions of when it comes to data aggregation is a full software like a service. You know, like and uh nobody's taking that service. Nobody's done that. I mean, I'm pretty sure somebody's trying to standardize, but like right now there is no one agreement of what standard going to look like. Every company doing something different. I do think like a completions has a potential to go, you know, like away from wits and all build something nicer, but can we agree as an industry? I don't know at this point because we all much smaller than we used to be 10 years ago, 15 years ago, smaller companies. We've been consolidating last couple years, which is okay. But uh what is part of dialogue between the companies from that side, you know what I mean? Uh we don't know. And uh this is why like from completions perspective, I feel like it's uh more of a service than you know, analytics at this point and the cost of it what we see today, it's way beyond the value it brings to you. You know, I mean, at the end of the day you just want to All you want is just visualize the data. Can I come back to uh Can I come back to like AI portion? Like majority of the customers today without paying anything just share teams like we do with you guys right now. And they just share screen and stream it for the That's a 2-week meeting. You can join that meeting anytime. I have my fracs I've streamed via GoToMeeting. [laughter] That's it. That doesn't solve the issue. Absolutely. You know what I mean? Like or you can go through the route of you bring another company, set up everything, all the workflow, pay thousands and thousands of dollars to have the same view on the chart, you know what I mean? And uh I think this industry we need to kind of come up to the not even like uh standard come up to easy and cheaper way to monitor. Uh otherwise it's just like it feels to me it's a service not a full software on that front. We definitely do work quite a bit of analytical portion of designs and cost optimization, everything else, but the whole streaming is just right now it's a little nightmare, you know, for any company. Yeah, and John I work in the Bobby, Bobby, are you on a background of completions or drilling? Uh none of them. I mean I I'm I mean in a way I mean now I mean I I have a math degree and then I started off at Conoco as a reservoir tech, but then John and I worked together at RDS where we were one of those companies that was, you know, capturing and streaming the frac data and we know the pain cuz, you know, we were one of the one of those first companies that was plugging into the serial port and then it's like, "Oh, someone switched the channel. Oh, we don't even know what this channel is." And in kind of kind of you know, way to standardize it and they're like, "Well, that works on Halliburton." Or this van with Halliburton, you know, but now I go to over to Liberty or I go over to, you know, Southwestern's frac fleet and like everything's different every time. So it's just Yeah, to your point, it's just much more of a you know, ad hoc thing and and again, I mean there's even like say the Coldbore's of the world and, you know, tremendous technology, but I mean even they have someone on site, you know, basically 24/7 and you know, they're standardizing, you know, as well as probably anyone. Um there's probably a handful more names out there, but it's um John and I know acutely the pain of the completions side and trying to give people any kind of good data. And then it just comes out to real time and what people call real time. Do they really need it? I mean Well Data Labs solved this problem probably, you know, 7, 8 years ago, you know, with post stage frac reports. So I mean are people actually making decisions mid frac or, you know, or do they just need last stages data to look at, you know, 10, 15, 20 minutes later, an hour later? I mean and is that sufficient? Um so there's all those kind of questions. And then you have That's one thing I've noticed with the completions is everybody doesn't seem like they're on the same sheet of music with what they want or what they see value in. If I go to a drilling rig and I say, "Hey, I got this software I'm building for you, give me ideas." I'll get 100 plus ideas within a month. But you go to completions, it's like that it's just not the same. It's completely different and you know, that's not really our our personal background. So that's why it's a little weird to me is to see the difference on what people see value in and completions from operator to operator. So there's so many knobs, right, on completions, right? And well, there's a lot of people involved, too. Or you know, it's similar to drilling, I guess, right? You've got reservoir and geology, and you've got all those other things. But I do think that just generally, it's just like you're sitting in front of a giant control board, and none of the knobs are labeled. And you're like, "Well, [ __ ] Which one do I turn first? Like, how do I How do I go about this?" Because there is just so much data, right? And it's not just about like you've got the frac design, and then you've got the chemistry of the frac design, and then you've got the sand loadings of the frac design, and then you've got how you do like the order of them, how you, you know, execute that. Then you've got the wellbore, how it's drilled, where it's landed, is it toe up or toe down, the stage spacing, the cluster phasing, the cluster distances. Like, there's so many things, right? But I do agree with you that we don't know what to do, like what to optimize, right? Which is why I think up to now, most people have optimized around the the operational efficiencies, right? The KPIs and how many stages per day, NPT, just like the general stuff. But you know, the dream state is on our line. That's part of the completion, but it's like, you know, now you have two separate operations that have to work together, you know, for the for the next stage to happen while the other one's going on. And it's just like, both of them can hold each other up. And again, like yeah. The dream state would be ways to understand what the frac is doing in real time, and ways to adjust it so that it's better than it could have been had we just followed the engineering schedule, right? Yeah. But no one does that. [laughter] Yeah, and then And then, then the problem there is like, maybe you could go off of a 3 6 12-month [ __ ] um but like, you're never going to get the real answer was this better, you know, necessarily, It was better because the rock was better over there. Yeah, right? Like, there's there's so many variables that make it hard to pinpoint whether or not what you actually did or changed was truly impactful, or was it just part of the variance and the noise of what we do. That's part of the puzzle we have to solve, huh? It seems it's a Plus, I noticed that I don't know very much about fracking, but it seems like it's also changing much more rapidly than drilling is, right? Drilling is drilling a hole. It's It's just changed a little bit of technology, drilling faster, different muds, different motors, different, you know, better, bigger rigs, bigger top drives, stuff like that, but I feel like completions is like drastically changing, right? And I I don't know much about like this What's it called? The simul fracs and stuff like that, where they're able to frac multiple uh uh well bores at one time, stuff like that. It's kind of crazy to me. But they're changing so fast. It's like they're kind of behind in terms of Well, that's the weird part, right? Is it's like, how do you innovate if you can't even agree on a a not even a format, but a naming convention. Just the names. Right? Like, no one agrees on a standard naming convention for a frac. No less how that should be transmitted and in what structure and in what format, how frequently. And so, like, how do you do any of that other cool stuff if you don't know, you know, if I call pressure and rate something different than you do, like, we're reading different maps, you know, like, that's part know why in our industry it's so hard to get people on board. Uh like, always one of the things that baffles me is, you know, AI driller's version of our system of record. You know, we have like fully automated reporting. And like, this is the future, right? Like, this is so obvious. Everybody knows it, everybody sees it, everybody's blown away. Company man doesn't have to do reports anymore. But you go and try to tell operator, "Hey, I got this for you." It's like, "Oh, well, we can't we can't do that right now." You know what I mean? It's just like man, it just seems so easy. Why can't we just go to this the the group as the IADC or whatever it is and I just change the standard to this like fully automated standard. You guys set a standard and go, you know what I mean? But, nope, people still going to they're going to Well, that's still keep writing by hand until until one day everybody's doing it and they're going to be like, "Man, I wish I did it earlier." You know? Yeah. That's the annoying part is it's like you you have all these conflict everyone has the same problem. But, then they think we all have to come to the exact same solution. And it's like even for me like for frac, I've been telling Bobby and I have been talking about this for years. It's like just publish your naming convention. Publish your data schema. Regardless of whether it's frac drilling if you're a software company, publish this stuff and let other people use it. Right? Like holding these things in silos and then letting people do their own things makes it harder for everyone, not just for you. Whereas if like whoever major E&P published, here is our frac data schema. And this is what we use. Maybe some of the service companies pick up on it and keep you it just like filters down, right? But, no one really does that or talks about it. They might have it in some documentation somewhere, but it's like because we don't talk about [ __ ] in our industry, [laughter] then no one learns from each other and then we all have to learn these mistakes at the same time instead of learning and progressing much faster. But, uh the good news John on the good side like there is a effort going on by industry. It's called OGDC. It's all in gas data quality meeting and it's led by um Zoom or ConocoPhillips and potentially Oh, Anadarko. Shout out to Zoom. potentially might be a good candidate for you guys to bring here just talk specifically on the what's going on. There is a couple efforts already been successfully tested. So, we kind of excited for you know, waiting for them like to agree on some of the those ones and that will be part of that testing as well. But hopefully, you know, like it would that that will become that what you exactly explaining like at least a standard how we do it. From the side because you know, everybody knows about this of completion work like you know, you have your wireline, you have your frac side or frac frac lines and you have like your drill outs, you know what I mean? But like nobody as you said agreed what hook one should mean or what tension should mean, you know, from that side. What pressure name, you know, I mean it's it's all a channel channel one, channel two and sometimes they change them in the middle of the job. For whatever reason like I I just feel like it depends on what stage it's on. Yeah, yeah. I just feel like a fighting wire is going to be better, you know, I mean like that day specifically. Mhm. So, just make everybody's like I think those might be I think there's two things, right? I think operator led coalitions, consortiums are way better than service company led ones because the service company led ones typically end up being led by one of the major service companies that have lots of incentive to force their agenda or whatever they want. Um but the other thing is it's like we've we know what a data standard does. Look at what drilling has done and what how much further ahead drilling is in all of this stuff than completions because it has been enabled by having this data standard. Yeah, and yet no one wants to agree on it. Yeah, but you know, if you think about it like look come back to my point like drilling did not change. When we went on conventionals, we drilled the same way with the same mud, with the same tools. It's just a little bit different formation and all we need to do our goal is to make it fast. So, all the drilling drilling efforts like worldwide globally in last 34 years kind of led to, you know, like operators get together what bits ML should look like was 1.31 version and then there is a some organization came in early like just to kind of take it more standard. But on completions we've been we started heavily completing in US probably back in 2014, so we kind of like 10 uh 15 10 years into it, you know what I mean? When we actually start heavily going through uh growing that industry overall. You know what I mean? So, this is why like I feel like we just need to wait another deep years and uh 5 years or 10 years it's going to all get better. I hope so. Marat, look at this guy. Sorry, Richard is uh currently uh updating my stickers on my cup to make sure that y'all are equally represented. I got to get that. [laughter] There it is. Well, that that was a very quick hour. Um do you want to try and do the speed round, Bobby? Yeah, let's just jump into it. Um So, yeah, so I guess this um so then Marat's more the dev side, but you can jump in if you have a favorite one too, Richard, but uh is your favorite database? I don't have one. Like uh for me uh I trust our team for those kind of decisions, you know what I mean? Like uh all I care about um cuz uh you know, my background is directional drilling, right? I mean uh and performance drilling. All right. All I care about um you know, like definitely I'll get much more intact, but if you take me from like you know, 10 years ago to now, all I learn is like things are much more basic than it looks initially. And uh you know, more sophisticated explain to the guys uh more better we increasing communication, you know? We we to kind of speak in a normal language. People think of laptop and server has no difference from laptop. This is a larger machine from that side. And I think what the my main goal is from customers is to make sure we go extremely fast faster than anybody could ever achieve. This probably jump into C++ and to like a very raw levels to see like what the possibility is. Uh but and in the same time like I don't care what's on background. And I don't want customers to care because when customers start talking language of like AWS, use this, use that. What's business value? What are we trying to solve? This is what we should as subject matter experts be focusing on. The technical things we do in the oil and gas is not new. Like it's already been done by millions of companies and hundreds of thousands of industries around the world. You know what I mean? Like we already behind. So ideas like we all we need to be much more crystal clear what do we need to solve and focus on those things and everything else is nuance. What infrastructure I'm running on? Like you know, what kind of like a data lake is not going to be bigger than you know, like Google's data lake right now what they use for, you know, Gemini and all that stuff. It's not going to be bigger than you know, like the GPT's database or data centers from that side. But I get like from like this client perspective like why would you care how ChatGPT is run because you will never have that big of an infrastructure. All you care about what the lead engineer wants Yeah. You know what I mean? And this is kind of like a what's the main the main goal of the CTO on that front and uh on my end is just you know, I represent users in a every dev discussion. You know, like I want to make sure we give it quick cuz as a user I want something quick. I want something I can use, I can click around, I can learn. And if we need to toss that feature later on away and build from scratch, let's go for it. Because ideas like uh I'm drilling right now. If it doesn't help me right now, I'm losing money. And this is mainly the goal what you know what I mean. I'm bringing to the table coming from offshore and conventionally I mean on this front. But definitely like all I'm learning now is like you know all the discussions and planning and everything else is just I try to stay separate myself from the trying to be even subscriptive on the what technology I've learned from everybody. You know what I mean. Like all I want to be just exactly clear what do I want from user perspective. And then the chain goes [clears throat] decides on everything else. Well, for the last one I'm going to go off script and make a announcement that we've been discussing, but we're soft announcing that our partnership with Collide where you guys are going to pull in via our API in through your platform so you guys can enable your customers to leverage the knowledge base of Collide directly in the platform and um [snorts] additional integrations. We don't We don't have all the details even lined out yet, but we're really excited about it and I'm excited about y'all's willingness to you know look at new and innovative things like what we're trying to do and you understand how it you know integration and and distribution of these things you know that network effect compounds across both sides of things and so we're really excited about it. We'll do a whole episode later on once everything is inked and formalized and integrated and we can actually talk about all the features and stuff, but I'm excited [snorts] about that for you guys. Yeah, I mean appreciate it man. Uh Oh, go ahead. Yeah, it's kind of like is cutting people off and he he blames his bootleg internet. He always does that. [clears throat] That's what he always does. Go go go go go. [laughter] No, go ahead man. We We are super excited to work with with Digital Wildcatters with the with the Kolide GPT and uh, you know, the funny thing when you asked about the AI, when I ask every single person who asked me, "Where's my AI at? Where is my AI at?" I always say, "Where's your AI?" and they always say ChatGPT. So, now we got Kolide GPT, by far and away the best GPT type service in the in the energy oil and gas industry. So, we're pretty proud to to work with you guys. Plus, you guys are awesome dudes, you know what I mean? So, it's kind of great synergy there. Yeah, we're excited, man. It's going to be a lot of fun. I hope it's just the beginning of more things to come. So, For sure. And this what Jim kind of from my from my end is kind of back to your Amazon question, like, you know, what company should be prepared in terms of infrastructure? So, the same for GPT, like, we have a lot of like enthusiast of drilling in the platform. Like, these are our main champions. They love things. They They enthusiast of technology at the same time. They like new features and thinking of how it can change the efforts. So, we're really excited to bring the your Kolide GPT to them, but they asked us multiple years ago. They said, like, "Look, Murad, when you're going to have this tool to, you know, like, when the I can ask questions and it gives me answers and you know, like, on that front. And of course, we're bringing it on drilling side when it comes to like recommendations of drilling, you know, like, alerts and over pulls, but I always told them, like, whenever somebody much smarter than us will come and build it and we'll just come with integration and partnership and, you know, like, in the cuz it's a whole focus, like, you know, you guys need to think through every single aspect of oil and gas, every single business unit needs and project and training this uh, the team you have is uh, extremely, like, I don't know I don't know how you guys pulling it all together doing all the podcasts at the same time, but uh, that's extremely exciting. So, I'm really uh, looking forward for our users to have, like, you know, our to work with collide GPT and share the feedback and everything else. So, that's uh uh that's going to happen very, very very soon and uh I'm I'm very very very excited as well as Richard and I think rest of the guys we already have some users waiting for it. So, uh checking on daily when it's coming. Awesome. [laughter] Well, perfect. Well, guys, Colin is literally knocking at the door here cuz we've got another podcast coming in, but thank you all so much for joining. Murad, thanks for staying up with us. Richard, thanks for sitting across from me and acting like we're not sitting in the same room. [laughter] Bobby, we'll see you in a little bit. If you guys like what we're doing, make sure you like and subscribe and leave a comment for these guys.