Subscribe
Share
Share
Embed
Welcome to Chuck Yates Got A Job with Chuck Yates. You've now found your dysfunctional life coach, the Investor Formerly known as Prominent Businessman Chuck Yates. What's not to learn from the self-proclaimed Galactic Viceroy, who was publicly canned from a prominent private equity firm, has had enough therapy to quote Brene Brown chapter and verse and spends most days embarrassing himself on Energy Finance Twitter as @Nimblephatty.
00;00;00;03 - 00;00;28;13
Unknown
The one thing I really wanted to talk about was just like how we've kind of, you know, every blind hog finds its acorn. But just discovering product market fit has been interesting. Take out whatever like I like. Yeah. The truffle beef that I need that clip please. I'm like, I thought we were already rolling, so I'm good.
00;00;28;13 - 00;00;53;25
Unknown
If that goes out there. Well. All right, all right, kick us off. Can't tell us about the email you wrote a year ago to VCs. Yeah. You know, I'm back out in Silicon Valley right now raising capital. And I was looking at some of my emails from exactly a year ago, you know, November, December 2024 and just insane to see how much our company has evolved in a year timeline.
00;00;53;27 - 00;01;09;16
Unknown
Seeing how I used to talk about our company and how we talk about it now, and I was sitting, with Canisius today because we're just talking about how much velocity we have. I'm like, yeah, dude. I'm like, how long have you been here? He's like seven months. There's like two years. Like just just given how much we've accomplished.
00;01;09;16 - 00;01;29;17
Unknown
And I was like, Holy shit. That's like going seven months. That means the rest of our tech team has only been here, you know, less. Much less. Yeah. Three months. And so, you know, today I just kind of wanted to give, state of the Union, not even a state of union, but just an update on on Callide and the progress that we made this year.
00;01;29;19 - 00;01;55;13
Unknown
And kind of what we have coming down the pipeline and, looking looking forward. And in 2026, I think that it's just such an interesting time. You know, we run a four deployed engineering model, which we can sit here and talk about, for a minute. You know, Palantir famously coined the term for deployed engineer. And essentially, you know, what it means is, you know, we've productized 80% of our AI technology.
00;01;55;13 - 00;02;18;04
Unknown
And then as we move into clients, there's 20% it's custom implementation and integration, and it's our forward deployed engineers that are sitting there and our clients taking our technology and applying it to our clients workflows. And so those PhDs, you know, they all have deep domain expertise from oil and gas. You know, they're geophysicists, completions engineers, frack engineers and just really understand the pain points.
00;02;18;04 - 00;02;41;03
Unknown
But I think what's so interesting, I think like the biggest development we've had over the last couple months is we hired this guy, Sam Texas, and I hired him for two reasons. One, his name is Sam Texas. Two, he has a mortgage, but he also just absolutely cracked out. Extremely smart engineer. I think Sam's probably one of the smartest people in the world right now when it comes to using AI tools and software engineering.
00;02;41;05 - 00;03;02;18
Unknown
And I think that AI coding fundamentally changes the for deployed engineering model. You know, when Palantir had it, you had two teams, you had an echo team, and you had a Delta team, and their echo team was their team of people with deep domain expertise. And they really kind of served and a customer success account manager function. And then their Delta team was their team of software engineers.
00;03;02;18 - 00;03;24;17
Unknown
They could sit there and prototype really quickly. Or AI tools are allowing us to prototype in a matter of hours if at. And so you really don't need this team of dedicated software engineers for prototyping. You can just have your team of, domain experts. They come from oil and gas sitting there. And it's not like we're vybe coding.
00;03;24;19 - 00;03;48;05
Unknown
This is the thing that I think it's really important to differentiate. I mean, this is production grade code. And, you know, we've really built out these systems. Canisius, you and I were talking about this morning of kind of building out this, this sandbox. And so like the overarching theme that I want to talk about here is how we're not only building an AI first product, but our company is AI first, because all these companies are trying to become AI first or AI native.
00;03;48;08 - 00;04;04;04
Unknown
It's like, what the hell does that mean? What does that even mean? But we were talking about this this morning, you know, kind of talk about like some of the bigger companies that we talk to as friends, like other tech companies that we're trying to help become more AI native and like the challenges that they know of. Yeah.
00;04;04;06 - 00;04;25;04
Unknown
Like, I, through just within our network, I've been approached by a couple of different companies who are multibillion dollar companies or, like, getting pretty close, and all of them have kind of been like, okay, can I can you tell us what you did, how Clyde is going through this process, and can you make us AI first?
00;04;25;04 - 00;04;55;11
Unknown
And my first question to them is always, what does AI first even mean to you? And I think all of these big companies are really struggling with the fact that they want to use all these tools. They read about all of that, but never understand how to really incorporate into their workflows. And one of my biggest learnings coming over to Callide was, I think if I didn't come out to Callide, it would have been the same thing for me because I was head of engineering for multi-billion dollar company and this was all the same situation.
00;04;55;14 - 00;05;28;25
Unknown
The thing that these companies have to really do is one have the right mindset, but they need to also get that expertise and experience from the outside and integrate that. Because if you're just trying to build this within the company, especially bigger companies, there's no way, right? And one of the other light bulbs that went off for me was that earlier when I came in and joined Clyde, and we were talking to a lot of these bigger companies and our clients and our AI first folks, I always felt that we were far behind because they would all talk about, oh, we have our custom model, we have fine tune.
00;05;28;25 - 00;05;59;11
Unknown
We were just talking about in the morning that somebody had said that they have a fine tuned model. We were like, okay, what does that even mean? Saying that they have a proprietary model and it's just a series of instructions to a general foundation model. And yeah, and now when we talk to these companies, we can go to such depth and really realize that all of these companies are simply just using out of the box very minimal ChatGPT prompts and all these other things to just go to pass the data to do these workflows, which is just never how they will scale.
00;05;59;13 - 00;06;18;07
Unknown
So a question I used to be scared off, which a lot of these companies are now getting scared off, is that they get compared to how are you better at the GPT right? Or copilot? We just can very definitively showcase those things. And whenever I talk to any of these big companies that are now approaching us, nobody's asking that question of us anymore.
00;06;18;07 - 00;06;37;07
Unknown
That's been the other. That's pretty soon. That's actually a progression that we talk about. Like you remember when we first started doing sales at the beginning of the year, and it was me and you convincing people that we weren't just to podcast operas? Okay. So, hey, that's how Chuck started every every sales pitch. Hey, look, we're not just, to podcast, bro.
00;06;37;11 - 00;07;00;07
Unknown
You know, we got some software developers, and. Yeah, it's just, like, convincing people that we weren't grifting selling. Selling it. Well, you know, you know what the to to take it out of the technology realm. And to give an analogy of this, if you think about kind of how software built up over time, you basically divided the enterprise into silos.
00;07;00;07 - 00;07;32;24
Unknown
You you created finance, you created human resources, you created operations, etc. and you if you're an existing player in any of those you drilled down into that silo, if you will. And the actual benefit of AI is to be able to pull all of that stuff together to automate those workflows and cut across there. And as simple as that sounds, that's a huge mindset change.
00;07;32;24 - 00;07;58;18
Unknown
If you're an application that does just safety, well, how does that integrate with finance, etc.? And I think that's part of the change that people are struggling with when they're thinking this through, even to make sure they're done specifically for oil and gas. You look at upstream, you know, you have drilling, completions, production, accounting, land, you know, whatever it may be, reservoir.
00;07;58;20 - 00;08;21;01
Unknown
And all of these are siloed on different tech stacks, data, you know, tribal knowledge within them. And so let me give you a perfect example. We were talking to an oil and gas company and potentially we're going to build this workflow for them. You know, if you're drilling a well, what are you incentivized to do? Throw it as fast as possible as cheap as possible.
00;08;21;01 - 00;09;00;03
Unknown
Right. Well, what's really interesting is if you come down and you drill and let's say you geo steer a little bit out, the the driller wants to get back in zone as quick as humanly possible, hit the target where if you would actually pause for call it 12 hours, you can figure out a more gradual way to get back in zone, because the kink in that in drilling that lateral right there is going to cost you half 1 million or $1 million over the lifetime of the, well, changing out pumps, those kinks, dug legs and when you drill, you don't give a shit about dog legs because that's not your problem.
00;09;00;09 - 00;09;19;05
Unknown
Exactly. The production then, but production is the the costs of those dog. What's funny is that was the mindset I had when I had rough night on rigs and drilled wells, and then when I started running expandable casing, I'd have to try to work expandable casing down these dog legs. And I'm just like, that's when I started seeing it because I was like, the drilling guys don't don't give a damn.
00;09;19;05 - 00;09;57;09
Unknown
You know, that's not their problem. But but an A and I a native drilling plan. That's an easy rule to follow. Minimize dog legs, you know, and and create incentives for that driller to minimize the dog legs. Yeah. One quick point that I wanted to kind of make on, on the AI transformation. I do think that one thing we did really well this year, and I think maybe Colin, you can give more examples of how we went through the journey was really change management, because I think as much as I this talk about AI, unless you don't have the right expertise and you're willing for your team to switch, that's the biggest competitive advantage we
00;09;57;09 - 00;10;18;03
Unknown
have, right? That our team fully switched. And it was that change management that we had to sit in in the middle of for a couple of months of just working that way, through which I imagine when we talk to our clients. Right. Like, Todd, you've probably seen that right? It's really the change management piece. I think McIntosh's done a great job leading that initiative.
00;10;18;03 - 00;10;37;26
Unknown
But I'll tell a story real quick about one of our clients because it's one of my it's one of my favorite clients and why they're a favorite client of mine is because there's like no hype around AI. I know hype the first time we went to salon, let's see. You know, like when we when our first product was like, well, search.
00;10;37;26 - 00;10;56;15
Unknown
You know, being able to search across well files, find information really quickly. So I was like, I don't think there's any value in that. And I just shut me down. I was like, okay, well other MPs say that like I think that there's value in it. And he's like, no, there's no value in that. And they call us back like two months later and we go there.
00;10;56;15 - 00;11;13;24
Unknown
Same thing. Just like not a lot of like it's a little bit like foreign territory for me because there's like no hype and like they weren't excited about it. And I'm like, man, we got to learn how to navigate this. There's an old guy in the room. He's like, could y'all help us automate our regulatory filings? As like, Todd's like, yeah, I think we can.
00;11;13;26 - 00;11;41;28
Unknown
And we start doing that for him. And now, two weeks ago, three weeks ago, they officially filed their their regulatory filings with the Texas Railroad Commission with a 100% acceptance rate using collide and in that, you know, it took an engineer six hours manually to do that to collide, call it 30s, whatever it was, it's going to save on probably a thousand engineering hours over over a year.
00;11;42;00 - 00;12;05;17
Unknown
And now they're all in and leaned in and to the point where, you know, they want to be communicating this to their investors of how they're using AI, which they should. Every company like every company, you know, I talked to the CMP the other day, they sold their MP big transaction. Him and the core team are starting up another amp private equity back, and he's like, yeah, I was at this.
00;12;05;17 - 00;12;21;27
Unknown
You know it. Happy hour. And guy comes up to me, he's like, what's your AI strategies? Like? I just rolled my eyes. He's I don't want to have fucking AI strategy. What are you talking about? So but then I went home. He's like I started thinking about he's like he's like, what is our AI strategy? He's like, and I listen to a podcast with you and Chuck.
00;12;22;00 - 00;12;42;02
Unknown
He's like, my gears just started turning about all the work that I do that is so tedious. And, and he's like, man, I bet we could automate this and automate that. And I think people are going to start to have this realization that one investor is going to be asked, like using AI isn't going to be optional.
00;12;42;04 - 00;12;57;04
Unknown
Like if you want to be fundable, you have to be thinking about how you're using AI. And I hate using the term of using AI because what the hell does that actually mean? You know, I, I hate like, you go in a company like, oh yeah, we're using AI. Like, what does that actually, what does that actually mean?
00;12;57;04 - 00;13;17;11
Unknown
Like here we we sell outcomes, right? And in workflows and however we make that sausage, we don't really care. You know, I don't care if it's AI or a group of, you know, you know, 50 people in a back room, you know, that they're sitting there crunching numbers, and clients, you know, they don't care. They just care about the output.
00;13;17;13 - 00;13;36;21
Unknown
But, you know, I think a couple of things on that change management process, and Todd has done a really good job with us. So we started saying up our FTE team this year probably, I don't know, three months, four months ago. Yeah, literally two months ago. Yeah. We were talking to like two years. It's like just we started almost two, three months ago.
00;13;36;21 - 00;14;02;02
Unknown
Not longer than that. Yeah. And you look at we have three PhDs right now. We have John, John is, you know, 15 years oil and gas, mechanical engineer, ba frack engineer for a long time. It turns out, you know, frack engineers that deal with a lot of downhole data are also just good and and data. And I put John, you know, John's been here with us for I think almost three years now.
00;14;02;04 - 00;14;27;05
Unknown
And it's so funny to me because John, we've been building rag since rag was invented. And I put John up against some of the best quote unquote, AI engineers that, that I've, I've talked to when it comes to data pipelines and things of that nature. And you have Nick. Nick, who's a geophysicist turned software engineer at Chevron, spent six years there building out internal applications, deploying them across teams.
00;14;27;05 - 00;15;01;00
Unknown
And you have Michael Michaels, 20 years completion engineer. He's at BHP, a couple of MPs at Halliburton, who's at korva. And like you get these guys like how much better has that made our job and our in our company. You know, you can start seeing the light bulb moment for our FTE team was actually taking an extending all the data pipelines, being able to take the application that's been built in, deliver something like tangible, like real outcomes in a matter of we'll call it weeks right now.
00;15;01;02 - 00;15;37;06
Unknown
And so it's funny to see kind of their light bulb go off. They can kind of build out, you know, entire workflows, entire, you know, taking all the frack data in the state, creating insights from that. And now seeing customers in that light bulb go off for customers. We're seeing those champions kind of switch into, oh, now I can go and talk to my accounting team and talk to my land team and, and figure out all these kind of typical, I would say manual processes, communication back and forth and actually deliver something that's not just on one data set, but across an organization.
00;15;37;06 - 00;15;56;28
Unknown
And that's been. Yeah. So my thing, my favorite thing about recruiting people for our FTE team, since it's all people, a deep oil and gas domain expertise, it's so easy to recruit them because you show them what Clyde can do and are like, holy shit. Like, I spent wasted so much of my life doing doing that. And so they just I buy into it.
00;15;57;01 - 00;16;24;15
Unknown
Well, I'm, I mean, I'm actually going to do a podcast probably next week with to the core question of what is our AI strategy? And I'm going to throw this out here and all three of y'all opine on this. And remember that I'm very delicate, so don't be mean. But no, because because I originally thought search was going to be the greatest thing on the planet.
00;16;24;15 - 00;16;42;18
Unknown
Because when I was at Kane, we would look at, in effect, huge data sets of wells. And why were eight of these 200 wells good. And if you could figure that out, then you could piece together the play. And that's what we did and that's how we made money. And I was like, man, I was going to do that on steroids.
00;16;42;20 - 00;17;11;13
Unknown
And I think that's right. But the way to get there and so if I were a CEO, might I strategy would be number one, automate my workflows, literally generate an ROI on every one of these things. Filing the regulatory filings, etc., go from 2000 hours a year to 20 minutes. Boom. Give me my ROI because that drives, change management.
00;17;11;15 - 00;17;40;19
Unknown
People get used to running the software. But the real home run, I think, is changing those engineers mindsets from automating a workflow to something way more creative and un engineering, like in. I'll give an example of I know if I can find this, signature on a log combined with this production profile, I know that grossly under stimulated that field, I can buy that field.
00;17;40;20 - 00;18;04;13
Unknown
PDP ten do refracts and make ten times my money that pattern recognition into the future on big things is what's going to allow oil and gas to be cool again. Yeah, well, like wild Cat or like Chuck, to your point, you and I under understand the value of search, but I don't think that I think that the value of search will be extremely valuable.
00;18;04;13 - 00;18;26;20
Unknown
Long tail in the future as you get all of your data in there. But then as people learn how to use these systems as a thought partner as well, and this is where that like that, that change management and like user behavior change comes. And this is like a material mindset shift. I'll tell you an example of me using grok, a couple weeks ago I woke up one Saturday morning.
00;18;26;20 - 00;18;46;06
Unknown
I was like, hey, I want to build the GPU cluster at our office. And I started asking, you know, I told grok what I wanted to use it for. We had some data. We're fine tuning a model, right now, the shameless plug that's already performing every foundation model, petroleum engineering test and task. And it's our first stab at it.
00;18;46;08 - 00;19;10;01
Unknown
And so I was reading, Jasmine. Jasmine is one of our AI engineers. Super, smart engineer, actually worked on ChatGPT reinforcement learning team while she was at Microsoft. And I read her paper that she wrote on, this first pass of fine tuning, and I was like, okay, we can do this up in the cloud using a 100 H100 GPUs as it.
00;19;10;01 - 00;19;35;21
Unknown
But I wonder if I just built our own GPUs at our office. What I could do. So I sat there on a Saturday morning. I woke up at like 6 a.m.. My two hours is just me talking back and forth with grok. Hey, I uploaded our paper. Here's our results. Here's how long it took us to train. Here's what the cost were, here's what we're trying to accomplish, and started giving me a list of, different GPUs, Nvidia GPUs.
00;19;35;23 - 00;19;53;01
Unknown
And it put me on to the Blackwell 6000. Because at first it was giving me kind of like consumer grade GPUs, and I was like, no, hey, I'm willing to spend some money is like, okay, if you want to spend some money, you should you should get this one. And anyways, I woke up that morning with the idea use grok as a thought partner.
00;19;53;08 - 00;20;15;19
Unknown
Had really good performance and cost analysis across a 101 hundreds and in our own GPUs. I went and spent 20 grand that night, bought the parts. Three days later we had the beginning of the cluster built here and what you just did, what you just did. The analogy I make to that is the beginning of the shale revolution.
00;20;15;22 - 00;20;39;25
Unknown
The successful engineers and the beginning of the shale revolution were actually not the great established engineers, because they were so rooted in convention that they just couldn't wrap their head around. Why should we use fine mesh sand to frack, you know, source rock? It's not going to work. But it was the people that said, screw it, let's try.
00;20;39;27 - 00;21;07;25
Unknown
And it's the people that say with data, screw it. Let's try running, search, finding correlations, figuring out is there actually causation to it, are going to be the people that really benefit from I give you I'll have the step function higher. Funny anecdotal story. I posted a picture so our GPU clusters called Spindletop had posted a picture of Spindletop on Twitter and someone from a major service companies like, it's like, oh yeah, we're speaking out doing this.
00;21;07;25 - 00;21;24;15
Unknown
But, you know, we're we're worried about constraints with the single socket CPU like you have on that Threadripper. We're looking at double socket and all of this. And I was like, yeah, the difference between me and use it. You're going to spec this out for six months when you're one of the biggest service companies. What does it cost you to go spend 50 K doing something?
00;21;24;15 - 00;21;49;05
Unknown
I just woke up and did it. And then you texted Jimmy and said, we need a surge protector. We got Jimmy Server over there building this thing. It was good. And but through that, the iterative learning that you get, it's led to me developing this thesis of, you know, me and Sam are talking about this insight the other day.
00;21;49;08 - 00;22;09;15
Unknown
The majority of of inference is going to happen on device. And so I'm actually a little bit bearish against this CapEx trade in AI, where you're spending billions and billions of dollars building these centralized data centers, because the majority of compute inference is actually going to happen on device. So you look at my Tesla, it's computing right there.
00;22;09;15 - 00;22;29;01
Unknown
You can't you can't have any latency when you're doing self self driving. It's got a GPU on board. It's going to happen on phones. It's going to happen on glasses. It's going to happen on your computer. And we're just getting smarter. On the inference layer, controlling our own compute and hosting our models on it, training our models on it.
00;22;29;03 - 00;22;48;19
Unknown
And those are the collective learnings that that allow us to get better and our engineering team to get better. And, you know, talking about our engineering team and our FTE team, I want to talk about this because our teams like, this is the thing I like talking about most, like, I think we've built a fucking stud team and I'm just incredibly proud of it.
00;22;48;21 - 00;23;11;25
Unknown
I got to Silicon Valley last year talking to a top tier fan like Colin. We love you. We think this is one of the biggest opportunities in AI, but you got to increase your software DNA. And to be fair, we even have a single full time software developer on. It's like me. Oh, you were in, but I wasn't even full time.
00;23;11;27 - 00;23;39;03
Unknown
They me like ten hours a week from Todd. It's John, it's Julie and Chuck, and, so very, very, very, very busy like I, I'm not mad about it. And then, you know, off the back of that, they didn't believe that I could go recruit the best AI talent and tech talent. And, you know, first we got Canisius and I had to convince Canisius to be stupid enough to leave his head of engineering job at a multibillion dollar platform to come here and build with us.
00;23;39;03 - 00;24;06;03
Unknown
And, he was crazy enough to do it. And then that starts compounding. You know, I wouldn't have been able to convince all of these engineers to come over if you don't have someone of Canisius caliber. But then that led to, you know, we have some great young talent and and Stephon and Jimmy and then bringing over Mark and Klay from, you know, this is this is what's great about someone like Canisius is Canisius has 20 plus years in the industry.
00;24;06;03 - 00;24;25;04
Unknown
And you get to leverage his network and bring in the best engineers that he knows. And so, you know, bringing those guys over and then our AI engineers, you know, John Slow, was seven years at AWS. He's an expert at orchestration, and was actually deployed in oil and gas companies. I actually worked at Devon straight out of school for computer science.
00;24;25;06 - 00;24;48;19
Unknown
I don't know if there's a smarter person at the intersection of AI and oil and gas than than John. On our team. Then you bring Jasmine and Jasmine. Is just a subject matter expert in reinforcement learning. Worked at Microsoft, as I mentioned earlier. And it's funny how the tone has changed from, hey, we don't think that you can recruit the best engineers to.
00;24;48;19 - 00;25;08;03
Unknown
I was in San Francisco two weeks ago, and they're like, how the hell are you convincing AI engineers from Seattle at Microsoft to come work in oil and gas? And I have a couple answers to that. One is people want to work on real problems. Do you, as an AI engineer, do you want to go work on a new note taking app, or do you want to go solve the world's energy crisis?
00;25;08;05 - 00;25;15;08
Unknown
Kind of kind of easy decision there, right? But then two is.
00;25;15;11 - 00;25;36;29
Unknown
High density talent. It's like a black hole. It's got this gravitational pull in the best of the best. Want to come work with the best of the best. And so this just compounds over time. And so, you know, now the reactions, you know, I have a lady coming in today that, I'm recruiting think super highly ever, you know, very season petroleum engineer.
00;25;37;02 - 00;25;55;24
Unknown
And her response is like, your team is fucking stacked in. And that's like the best compliment I could ever get. And I think it's fair to say, right. Like even Sam on our team. Right. And I got to mention Sam. Yeah. Sam, he has an interesting earlier. Yeah. He there's a mixture of Chuck job contests, and that's how we found him.
00;25;55;24 - 00;26;16;20
Unknown
Yeah. So sorry to interrupt you. Can we tell this story real quick? Because I want everyone to understand the power of content. Sorry to interrupt you, but this is important. You know, this is why I wanted to kind of plug in, Sam, because I think it's powerful to talk about. You know, Sam, lives over in Sweden. Actually grew up in Houston, but he's been in Sweden for 11 or 12 years now.
00;26;16;22 - 00;26;46;20
Unknown
Has he's the CTO of, health care startup. He was at Candy crush, managing site reliability teams, which, when you're eye candy crush, that's a real. Yeah, that's a that's a real thing. Yeah, that's real stuff. Yeah. Anyways, I get a message from Chuck one day is like, hey, this guy reached out to me on the pod, and, you know, I got on Twitter, I think any time Chuck sends me something, I go, it's like, you know, sometimes I pay attention to it, sometimes I don't, but I click on his LinkedIn profile and I'm like, damn, this guy actually seems legit.
00;26;46;20 - 00;27;08;21
Unknown
And I get on a phone call with him, and me and Sam just hit it off and and vibed. And anyways, I flew him over here and talked to him and anyways, fast forward nearly a year later and we're talking about our team and telekinesis. I'm like, man, this this guy, he was researching oil and gas stocks and looking for oil and gas information.
00;27;08;21 - 00;27;32;07
Unknown
And he came across Chuck's podcast, and this is in the early days of Clyde. And we start talking about AI and, you know, Sam just got it. He's like, man, you got the content in the community and in the AI. And anyways, Sam has been a huge addition to our team, and that came from Chuck's oil and gas podcasts of him telling whatever stupid jokes he's done.
00;27;32;07 - 00;28;05;20
Unknown
God knows what. But the content is amazing, both from recruiting side to customers. Like almost every time we go into a conversation with the customer, it's like, oh, I saw Colin or Chuck on a podcast. And that got me thinking. And here we are talking about AI and what we can do and like deliver outcomes. It's amazing to see the response, dude, I need you to talk about this a little bit because the shape of this, you know, to preface this, Todd, been in energy 20 years, but that's like 20 years.
00;28;05;22 - 00;28;40;12
Unknown
Todd's done everything from, you know, working at Chevron. He founded and sold a, startup in an energy, you know, he's one of the leading, subject matter experts on CSAs and built a consulting business around that. All that to say is, Todd is seen a lot like, I think if you draw a Venn diagram of oil and gas domain expertise, software experience, and then like commercial startup experience, like, Todd may be the only guy that kind of fits across all three of those.
00;28;40;12 - 00;29;11;08
Unknown
So you've seen a lot in Todd. I know when you came over here, like you just saw that our distribution list was just amazing. Like you start talking to, any customer, any, you know, essentially any oil and gas. I'll throw in almost any oil field service company, and we probably have at least 2 to 3 contacts and relationships, not just like people we know, but contacts at those companies where we can go in and at least have that initial conversation, which is so unique.
00;29;11;08 - 00;29;35;11
Unknown
And you throw in, obviously, Chuck's relationships, your relationships, Jonathan's, John's, and now all of a sudden you're talking about distribution across upstream, midstream, downstream. And it's like, well, it's the opportunity suit. I remember, it's like a couple months ago. Chuck's like, yeah, man, we're in this sales meeting and CEO barges in. You know, it's one of the largest private companies CEO barges in.
00;29;35;11 - 00;29;57;15
Unknown
It's like, oh my gosh, Chuck, to listen to your podcast I take a picture and Todd's like, dude, I was there. I saw it happen. I witnessed the whole thing. This great stop meeting. So Chuck Todd let me get that. We get to something that's great. That was that was very Dave Grohl ish. Yeah. Those hands go. But, you know, it's funny about that.
00;29;57;15 - 00;30;22;11
Unknown
Is, you know, we chew glass building this like, you know, it's hard is bootstrapping a media business. And media businesses are actually really shitty to operate because it's just like I have a lot of respect for people that that grow and operate media businesses because they're challenging. It's not a very repeatable thing. And but the thesis was, is that, hey, you could build community with content.
00;30;22;13 - 00;30;45;10
Unknown
And I thought content was the highest form of leverage of our generation. And so, hey, we'll use podcasts. Will you use events to build community and then use community as your distribution? You know, so funny. If you ever get on Twitter there's two camps and tech kids team product or team distribution. And some people think that, you build a world class product, people come use it other camp things.
00;30;45;10 - 00;31;06;05
Unknown
And this is there's like a running joke. It's like second time founders or first time founders focus on product, second time founders focus on distribution. We built our company backwards. We spent five years building distribution and then focused on building world class product to go distribute. But it's been so cool to just see that vision come to life.
00;31;06;05 - 00;31;26;06
Unknown
And I think that really hasn't even been until like the last two months that I'm like, oh shit, this is like, this is what I wrote about in that original paper years ago. You know, we had a team on site and we're sitting there. Think it was John Slow. And Jasmine's like, we had an idea in real time.
00;31;26;06 - 00;31;40;21
Unknown
And then John Slow and Jasmine are going back and forth. I just sat back and I was like, I feel so grateful to be sitting here with these people also. That's part of being an entrepreneur is, hey, if I want to learn about reinforcement learning, I'm just going to go hire one of the smartest people in the world on it.
00;31;40;21 - 00;32;01;28
Unknown
And I'm just going to learn everything about reinforcement learning from them. But the see that in Chuck? I think he even had that that, feedback was sitting there on that team meeting and you like, sat back, you're like, look at this. This is real. It used to be all sitting around scratching goes, oh, I say the comment on the podcast is that enterprise software company.
00;32;01;28 - 00;32;32;12
Unknown
Now what's let's keep it to those laughing. We use, granola. I know it's, for for all of our notetaking here. And they did this granola crunch, which is like Spotify rap where it gives you a year end, recap and all of ours had some comments about Chuck. It's like Chuck begrudgingly talking about his new role as AI salesman, or it's like Chuck's most likely to sell a feature that doesn't exist or whatever.
00;32;32;15 - 00;32;54;25
Unknown
But yeah, I think what's really exciting I was talking to Confucius about this morning is I think 2026 is the year of scaling for us. You know, when we came into this year, there are a couple things that we really had to prove out. You know, the first was like we talked about increasing our software DNA. Hey, we're a group of smart, scrappy oil and gas people, but we really need to get enterprise software DNA in here.
00;32;54;25 - 00;33;19;27
Unknown
Boom. We went and knocked that out of the park. Then is building the platform and the technology. Confucius has a chart that he likes to show every time he pulls it out. I think he's talking shit, but he's like, this is our architecture. When I joined, this is what our architecture is now he flexes. The funny thing about that is when he pulls out the before photo, I actually understand that the I don't understand.
00;33;20;04 - 00;33;43;05
Unknown
So you must have done something. And where we're at now still like where we're going to be at the end of this month. I mean we are tax. Just first off, the rate, the the velocity that we have and the rate of change here is absolutely insane. I mean, it'll rip your face off and our platform is going to be at the end of December is just, absolutely amazing.
00;33;43;05 - 00;34;02;01
Unknown
All the things that we're doing with our back end data pipelines and orchestration and fine tune models. And so that's what a lot of this year was about, was building up world class team and getting our platform to where it's scalable. I'll say the third thing was coming into this year was looking at the community platform on collide.
00;34;02;07 - 00;34;23;13
Unknown
You know, I got comments from top tier VC last year and they're like, Colin, is this community, is it is it real? Like we get on there and there's not a lot of engagement. And pretty much my response is, okay, I know that's what it looks like, but you have to trust me, bro. Like, I know this. Like I know the industry needs this, but that doesn't that doesn't answer.
00;34;23;15 - 00;34;42;04
Unknown
That's not a good answer to VCs. Right. And so we came into this year and it's like, hey, we really have to prove out. Like we know as oil and gas experts that the industry needs us and that the community needs it and that it's really important. And we need to prove that out. And we set we set some goals, to define if we prove that out or not.
00;34;42;04 - 00;35;04;01
Unknown
And we did it and we crushed it. And now I just see how much value the community is getting from it and how much it's growing. And it makes me really happy. And, you know, to give Todd some credit here, I was in an engineering firm and I was telling him about collide, tell him about the community. And at the end of it, you know, this is a room of like 15 engineers in there.
00;35;04;01 - 00;35;16;29
Unknown
And this guy's like, he's like, call. And he's like, you're talking about the community. And so I guess you're talking about it. I went and logged on. He's like, he's like, I got to tell you, he's like, I, you know, I joined a year and a half ago and you're like, pause. I was like, hey, dude, I know it was shitty.
00;35;16;29 - 00;35;43;08
Unknown
He's like, yeah, it wasn't good. And the technology sucked. You know, if we'd built it MVP off of no code using bubble God, dude, you go tell that to John. John like gets nam flashbacks and there was no engagement. And I'm very thankful for the people that were in the community at that time that just really believed in the vision and got it and were trying to like, push through the shitty technology.
00;35;43;11 - 00;36;09;12
Unknown
And this is when we got Todd involved. This is probably like 18 months ago. Yeah, probably. Yeah, yeah, yeah. And Todd started working with us as a contractor. Him and a few other engineers and, built our collide community full stack. And because that was a big commitment for us at that time, like putting as much money as we had to put into building that.
00;36;09;14 - 00;36;32;21
Unknown
And we didn't plan on it like there was no plan of monetizing this. It was just the industry needs this, and we're going to do it. And we think that this is core to our our community and our platform. And now you look at it and building network platforms is incredibly hard. It's incredibly hard. You always have chicken and egg problems, you know, how do you get people on there?
00;36;32;21 - 00;36;54;10
Unknown
How do you get them engaged? And what was interesting is we had people that wanted to engage, but the technology sucked and weren't allowed to engage. And Todd came in here and built up that platform, got it, got it to where it was usable. And now we've iterated, iterated, iterated to where the technology is good now. And you look at it, you go on there and people are having great conversations.
00;36;54;10 - 00;37;18;00
Unknown
People are getting value out of it. I got an email the other day. Someone was like, hey, someone hooked me up with a, fuel supplier. Can I send them 500W to, to, pay them for their time? Like, people are using the points and watch that they're getting and collide as a form of currency and that, you know, I'll tell just, I don't really tell this very often.
00;37;18;00 - 00;37;39;09
Unknown
Boat do. We launched Cloud Community three times. I think the first time was in 2021. Used a white label. That circle. I paid $8,000 a year for it, and I got people to sign up for it and say, okay, cool, people want this. But again, the app and the technology was shit. It was white label. You couldn't change anything either.
00;37;39;11 - 00;37;59;05
Unknown
And then it's like, okay, well, we got enough here, let's build our own app. And so we build it with bubble and fuck dude, just hindsight, I wish we wouldn't have done that. But I remember, like, there's videos of me launching an energy tech night and or we launched it first at horseshoe Bay event that we did. Then we launched it like a year later.
00;37;59;05 - 00;38;26;08
Unknown
We launched it three times before it took off. You know how stupid you have to be destroyed it like something over a span of 3 or 4 years before it gains traction. And anyways, I'm just in and, you know, shout out to Crystal and Julian, our team for cultivating that community, and Sydney and Misty. I can take so many parts and components to get to where we're at now in.
00;38;26;08 - 00;38;43;11
Unknown
I think 2026 is really the year of a scaling. You know, we had the platform where we're enterprise ready on our AI to go deploy this into companies. We got good technology on the community side. You know, we're leveraging the community. What's so interesting is you have all these big AI companies like scaling AI that do data labeling.
00;38;43;11 - 00;39;02;20
Unknown
Can they sell this to the large model providers? When I go recruit AI engineers from those companies, they see our community like, man, you have all this community of subject matter experts from oil and gas. You could be using them to for data labeling and QA, QC feedback for your reinforcement learning. So, you know, there's a selfish reasoning.
00;39;02;20 - 00;39;20;01
Unknown
Two for building this this community app. You know, not only does the industry need it and it's going to help a lot of people, but we're also going to be able to leverage it. I've got an AI engineer that we're going to hire in Q1. I recruited him off of Callide. I sent him a message on Callide. And so we get we get something from it as well.
00;39;20;04 - 00;39;43;14
Unknown
But it's just so cool to see all of these things coming together. And 2026 is really going to be the year that we scale. And, and just, just add some color to that and correct me where I'm wrong, but I think and this is dynamic, it's changing every minute. I think we have ten client or 12 clients today doing 17 different things with Collide enterprise AI software.
00;39;43;14 - 00;40;08;21
Unknown
And I bet by the end of the first quarter, we're at 25 or 30 clients doing 35, 40, 45 different things with, with Collide enterprise. And so that's that, that to me is ultimately we can talk about upgrading software, DNA and all this. So that's kind of the proof's in the pudding. And you that's like we rewind a little bit.
00;40;08;26 - 00;40;29;24
Unknown
Maybe this summer we delivered 1 or 2 packs a month, let's call it that. December. We're probably going to deliver between 8 and 10 proof of concepts where we can actually get customer validation on that. And now you start going into Q1, it's like what that what can that number be? Is that it's not eight and it's maybe close to 20.
00;40;29;26 - 00;40;47;02
Unknown
Like we're going to get it a velocity. That's just the industry has. And one of the things I'm most proud of is we built the technology to scale up and down the market. Like, I'm not interested in just building enterprise grade software for the biggest clients out there. It's the mom and pops that need AI more than anyone.
00;40;47;02 - 00;41;09;10
Unknown
I mean, they're getting squeezed whether you're an MP office. And guess what? Silicon Valley's not coming to see them. They don't even know they exist. Silicon Valley doesn't even know the big companies exist. We have a mom and pop company, I think two engineers, thousand wells, and we're building workflows that are saving them hundreds of thousands of dollars.
00;41;09;12 - 00;41;35;13
Unknown
And that means everything. If you can put that money back to their bottom line. And so that's a really big thing to me, is that we've been able to pull the most important thing there is, if you ask them right now, they'll tell you. Man five handle on oil price. Not sure my family's eating tomorrow, but if we, if we go chop 300, $400,000 out of their cost structure, that's a nice living.
00;41;35;13 - 00;41;55;17
Unknown
Which this actually gets me to a point is the most pressing issue that the world faces right now is an energy crisis. Over a billion people live in energy poverty. We know this in the industry. We talk about it all the time. But now people are subscribing to the belief that, oh, energy demand is going to continue to increase.
00;41;55;19 - 00;42;14;06
Unknown
I've been talking about this for years. You look at every energy demand forecast that's out there. It's wrong. You know, they'd have us plateauing by 2050. So you have all these secular drivers, you have high performance computing, which mind y'all, we were doing bitcoin mining content and events before the data center AI rush. We knew this was coming.
00;42;14;06 - 00;42;39;02
Unknown
So we talked about it all the time. You have electrification, you have emerging nations and then you have space. Colonization, which is space colonization is growing exponentially. You know, the payload, the cost per pound of or kilogram of payload is, is just going down exponentially. And so you look at all these drivers, energy demand has the potential to scale infinitely.
00;42;39;02 - 00;43;10;23
Unknown
And that's a good thing for society. And it's our community and our AI that is going to enable that. And you talk about increasing the margin and allowing these oil and gas companies to operate at cheaper prices. It actually brings down your your breakeven cost of, hey, maybe that breakeven cost, true, breakeven cost was 60, $65 a barrel, but now it's $40 a barrel, maybe $35 a barrel, because we've we've changed that cost structure and lower price spurs more demand, you know.
00;43;10;23 - 00;43;35;13
Unknown
So yeah. Yeah, yeah I honestly believe what we're building here has the potential to change macro economics of of energy. And I think is on par with what the shale revolution did I believe that I don't I'm not just talking my own buck here. Yeah 100%. Look dude, you were there. You were there, I was there. Yeah, I was a fucking angry 19 year old kid on West Texas.
00;43;35;13 - 00;44;20;24
Unknown
Like, fuck yeah, dude, were dethroning Saudi. Yeah, yeah. Chuck cutting the checks. And he had cracked out kids like me drilling the wells out there. And that changed, that changed the global stage. And I believe that this technology is on par, to do that as well. And, you know, it's about the people on the industry that are like, I think another thing here is the people at the companies that we're working with, you know, think of of I don't want to name any names, but just unbelievable forward thinking engineers that get it and see what the what the future potential of this technology is going to allow them to do and allow
00;44;20;24 - 00;44;44;09
Unknown
the way companies are doing is really grateful that they're those types of people, and the industry. And I think that, you know, we'll look back on this ten years from now and, and see it is on par to, to what shell did, in the transition from conventional unconventional, I, I do think that it's that material. Let's do this since we probably are going to get yelled at by Crystal on time.
00;44;44;11 - 00;44;54;23
Unknown
Kind of the exit question, if you will. What do we need to do next year? Scale. That's the one thing.
00;44;54;26 - 00;45;20;05
Unknown
Brevity of words. Go ahead, wash that, flush that out. So yeah. So how are we going to do that. And somebody out in the community that potentially wants to call and talk AI software, what are they going to get. Yeah, I think as Colin mentioned previously that we did invest a lot of this year just hiring really good talent, setting the foundation of our technology and also figuring out distribution.
00;45;20;05 - 00;45;41;13
Unknown
Right. You guys already had a lot of distribution. That was one of the reasons I joined. But now we are also with the right cross-section of having those champions who have really turned the corner. So I think for us now, it's really going to be how quickly can we produce the POCs and how quickly can we then put them into production through production.
00;45;41;15 - 00;46;05;02
Unknown
And I think that cycle time, I really do believe is just going to continue to shrink because of us truly being an AI first community. And we've had these conversations and even if happy to actually start a series, even on the community. Right, like just posting more about these things or if people want to kind of call in and happy to kind of set that time, but which I think that's going to be our biggest unlock as well.
00;46;05;02 - 00;46;44;24
Unknown
Right? Because we can get to production, get POCs fairly quickly, and really start going from 2 or 3 clients to hundreds of clients and applications within the next year. Yeah, yeah. And I would kind of extend that maybe a different slant on it, but kind of scaling the value so that as we start thinking about the pieces that we're delivering, then proving out those outcomes and seeing exactly what value we can deliver for each customer and each kind of opportunity, that, to me, is where you start looking at these, larger MPs that have billion dollar line items and where we can go and chip away at that.
00;46;44;27 - 00;47;12;27
Unknown
That's exciting. Right? We can basically have a forward deployed engineering team have the backing of true software engineers and actually deliver not just incremental value. We're talking like taking chunks out of the financial statement to see true value, delivered on epoch. Can you just said it with brevity of scaling. But I think the scaling actually has probably three facets to it.
00;47;12;29 - 00;47;35;05
Unknown
One is from a technical engineering perspective, two is from a commercialization sales perspective, and the third is from a financing, perspective. And so, you know, we've talked about it already here, but made a huge capital investment in our team this year, getting world class talent to build our platform where it's ready to scale. Boom, we're at that.
00;47;35;05 - 00;47;57;15
Unknown
We're at that point, you know, if anyone's listening to this, where we're going to grow on our team is our forward deployed engineering team. We're going to be constantly looking for the smartest people from oil and gas and energy as a whole. Bonus points if you have some Python and software development experience. I'm actually thinking with our system that we're building that you actually don't need a ton of experience there.
00;47;57;15 - 00;48;19;02
Unknown
Maybe not at all, but bonus points if you have a that's where we're going to really look for people. And, you know, going from where it took us, you know, months to do a POC towards taking us a couple of weeks. The second one, sales and commercialization. You know, this industry, it's hard getting your first client.
00;48;19;04 - 00;48;36;02
Unknown
And once you get your first client, the second one comes easier, third comes easier. You know, we're deployed in multiple companies now and that pipeline is huge. And so you know, what's funny is, is like we started getting ahead of sales. And I always use Chuck's analogy of like a water balloon. Let's like you push here and pops out over here.
00;48;36;03 - 00;48;55;23
Unknown
Then you go push it over here. And so you always have this tension between sales in engineering and deliverability. But you know, we're doing this really quickly. Just an anecdotal story of I want people that are listening to understand where the world's going to of our engineers. So Nick for deployed engineer, he's driving in his Tesla and it's self-driving.
00;48;55;23 - 00;49;16;13
Unknown
Okay. So the AI is driving him. He's got his laptop pulled out. He's got Sam on the phone with him on a zoom call from around the world in Sweden. And they took one of our prototypes and productionize it in the time that it took Nick to drive across Houston. So they actually Sam measured the steps of doing this.
00;49;16;13 - 00;49;45;06
Unknown
He said it took from, Beltway 8 to 290 to get this part done, and from 290 to highway six to get this part done. This is the future. Like this is this is the future, and this is how quickly AI allows you to work and how it enables you. Think about telling someone two years ago, oh yeah, I was in my full self-driving car on my laptop with someone around the world, and we used AI to develop a production grade app in 30 minutes.
00;49;45;08 - 00;50;14;06
Unknown
Like what? What the fuck? Crazy, crazy time that we're living in right now. The third part is the financial aspect. We've been very fortunate. You know, we've raised $14 million to date for venture capital backed. $11 million of that has come from angel investors. You have, you know, Chuck here. I don't know if I've ever given Chuck credit on on air, but, you know, someday I'll write a book and I'll talk about, you know, you know, Chuck changed my life and probably saved my life.
00;50;14;06 - 00;50;36;00
Unknown
Chuck was cutting 500 K when there was no technology. There was no business made. The screwed up thing is, when I let you have a crawfish boil at my house. 500,000 easy times. I didn't care about the checks that I got. He's like, you guys barfed on my third story. The Denver podcasters. And you know who you are staying up on my third floor.
00;50;36;01 - 00;50;55;28
Unknown
Yeah. And. And you know, we have just, an amazing investor base of the who's who of oil and gas and, just very fortunate because when you're a startup in Houston, Silicon Valley is not looking at you, especially when you're in oil and gas. Two years ago, no one gave a shit about oil on the gas. Now they're beating down our door.
00;50;56;00 - 00;51;13;16
Unknown
But two years ago, no one, no one cared. And so I've been very fortunate that we've had an investor base of angels that have allowed us to get to the point where we're at now, but now we're scaling, and this is a capital intensive platform that we're building. You know, building enterprise software is extremely challenging. We can't be at 80% on these systems.
00;51;13;16 - 00;51;34;07
Unknown
We have to be at 100%, especially in an industry where you're making capital decisions, you know, in the hundreds of millions, billions of dollars based on, you know, decisions that our systems are going to be making, or you make a decision that could potentially kill five people on a rig. This is serious, serious work that we're doing, and we take it very seriously.
00;51;34;07 - 00;51;54;12
Unknown
And it's a capital intensive technology and platform to build. And so, you know, we're going out to Silicon Valley now. We're going to go raise, $25 million in our series A and, you know, we'll probably raise $100 million over the next, couple of years, which is funny because you told on gas people this joke on $1 million isn't shit in line.
00;51;54;14 - 00;52;14;27
Unknown
It's a lot in the in the tech world. And, we're at that. We're at that point now, you know, Mercury fund invested in us, nearly a year ago and got us, to that point where we're institutional grade now and we have the ability to go to go raise those dollars. And so when Canisius says scale, that's really how I look at it.
00;52;14;27 - 00;52;34;09
Unknown
I look at it from the technology perspective, from the commercialization perspective, and from the financing perspective. And that's what this upcoming year about for us. So I think this upcoming year is going to be about is, you know, three years ago we would talk about AI and I would whip up some wild story just to get attention, right?
00;52;34;09 - 00;52;56;00
Unknown
Yeah. To lo and behold, some of that actually does come true. And I think now over the last course of the year, I am now taken seriously on we can automate workflows. I think next year AI will actually be taken seriously on collide is going to be the agent in charge of all your agents within the oil and gas business.
00;52;56;00 - 00;53;19;16
Unknown
And so that collide window pane is going to be how you, the oil and gas person, in effect directs and I'll say 100%, but let's say 80%, 80% of your workflows on a daily basis, no collide will be the operator. And I'll be taken seriously when I said it. Now, now I get a, an eye roll or what are you talking about?
00;53;19;16 - 00;53;39;18
Unknown
But I think that's actually good. I get there's a large company that we're talking to that does billions and billions of dollars of revenue, and they articulated that without us saying it. And so I think that the industry's already already getting to that point. But yeah, I agree with you that where last year we were having to convince people that we weren't podcast bro.
00;53;39;21 - 00;54;02;28
Unknown
Okay. So I had to convince people of that either. So yeah, but you know, one thing just kind of in this out, you know, I'm just really thankful for, you know, our team, for our community, for our investors. You know, it takes a lot of, different components and, you know, yesterday alone and slack, I shared three posts that I saw from LinkedIn.
00;54;02;28 - 00;54;22;28
Unknown
People talking about collide, and I could not be more grateful or stoked to have people that that we have in our community supporting this. And so, you know, our first core value at, at collide is community over everything. And, you know, just really thankful that we get the opportunity to work with, the amazing people that we do.
00;54;22;28 - 00;54;35;29
Unknown
So, with that said, you know, 2026 is going to be a big year for us, and we're running a thousand miles an hour. And, you know, I'm sure the update that we have in a few months will make this one look pretty antiquated and old. So let's get after it.