Welcome to uh Energy Bites episode 5. Got uh Matt Harman and Josh Groves from Pod 2 with us. Um no, just excited to have you guys on. Yeah, we're excited too. Thanks, Bobby. Got it. Happy to be here. So uh glad we could snag you while you're [laughter] Yeah, while you're in Yeah. Josh is Well, I'm just let him introduce himself. He's in from not just Canada, but Newf Finland, which is like a whole another world. So, [laughter] you know, happy to have you on while you're here. So, um, but yeah, if you guys want to maybe just introduce yourselves real quick, you know, who you are, where you're from, you know, how you got here, and, uh, go from there. Yeah, go ahead. Me first. Okay. Yeah. So, my name's Matt Herman. Uh, right now I'm the founder of Pod 2. You know, I was born and raised, you know, Texas, San Antonio, Amarillo, played baseball growing up, and so I wasn't good enough to go to a school that had an engineering program, so I did math. Um but then ended up joining Chesapeake Energy in 2011 not knowing what an oil was. Um an oil [laughter] I mean like your background is extremely similar to mine. So yeah. And so like I I literally got the job because I was spamming Oklahoma City with resumes to try to pay for my wife's engagement ring um just for like a summer job. Like oh here's a career by accident. Um got in there you know took six months to realize that I hated working for a big company. Um, but we were a customer of inner sights and I was running field development plans for the Udica for Chesapeake and you know liked it you know took to it and got had a good relationship with the inner sight guys at the time. Um, so joined them I think I was employee number like 20. Um, Josh joined Insight at like the same time. So that was like 2013. Uh, we grew up through Insight and as it got bought and bought and bought um started pod 2 in 2019. Uh, so I focus more on consulting. Um, that's where it started and then when Josh joined in 21, that's when kind of built out the the tech side of the business. Sweet. Yeah. And and so my background, um, I studied mechanical engineering in in Newfoundland. Um, we didn't have a petroleum engineering program, so I ended up taking a couple courses on top of the mechanical just to get used to that. Um, so then I worked with a couple operators, smaller operators in Newfaland. And then um afterwards I I was looking around and I there weren't a lot of uh job opportunities in Newf Finland. So I started looking around at other places like Alberta and Texas and I figured I might as well go right to Houston and and experience that whole side of the industry. So that's what I did. So I moved down and that's when I joined Insight and that was back in 2013 like Matt said. Um, so we're really small company, so working at a startup was really exciting and enjoyed that side of it. Um, and then when Matt went out on his own, I was pretty eager to join him and then that's when I joined him in 2021. So I I'm running the tech side of pod 2. Cool. So tell now that we've kind of established where you guys are at, tell us a little bit more about kind of pod two and what it is because y'all are kind of do both sides of it, so to speak. Yeah. [snorts] So the the underlying purpose to the company is to remove suffering for people at work. Um and I think the way that we do that is is in various like there's various ways that we do that. But the um one of the things that I think is a little bit different is that there's you know consulting firms and companies that focus on you know nothing but profit and you know helping helping others you know make more money, save money, all of that stuff. And they don't think about the people at all. And then on the flip side of that, there's, you know, culture consultants that think money is a bad word and just do woo woo kumbaya [ __ ] Um, and we kind of want to find the harmony between like between the two where, you know, doing the right thing for your people, um, really does lead to a better business outcome, you know, and more sustainable business, you know, better for everybody. Um, so the way that I do that is through consulting. You know, I hate the word management consulting, but that's kind of what I do. um a lot of process improvement. So helping companies figure out, you know, how a process is bad. Um how it should look in the future, make decisions, get people aligned, all that kind of stuff. Um also do a lot of like leadership development and culture, like team building, not from a trust fall standpoint, but from a like how do we actually be effective and get stuff done standpoint? So is that primarily like what's y'all's kind of targets? Like is there a specific y'all are looking at oil and gas, energy, anybody? And then on from like the kind of technical operations piece, is it strictly like looking at my tech operations or my data operations, whatever you want to call it? Um, or is it just more like all-encompassing kind of? The consulting side is definitely all-encompassing. like [clears throat] most of the almost all of our business is oil and gas, but then the the leadership development stuff, the team development stuff, we've done that kind of work with startups, we've done it with, you know, other consulting firms, which is kind of funny. Like consulting firm brings in a consulting. [laughter] Um that was actually like some really good work. Uh I was proud of that one. Um but yeah, it it's pretty all-encompassing. So, usually, you know, we'll we'll get to the root of a problem and say, you know, your forecasting is biased or, you know, you have competition that's unhealthy between assets or this whole process just takes too long. Um, that kind of stuff. And then we're really good at breaking that down into symptoms and causes. Um, breaking that down into, you know, there's multiple projects that usually come out of something like that. It could be data project, could be a software tool need, could be a process change, people change, like all that kind of stuff. Cool. No, that's I'm glad you're you're not just approaching the well this data needs to go over here like the integration or you know just like true consulting kind of dev stuff because you can have the best tools in the world, but if the people that are in the field that need to be using them aren't using them or they're not designed the way that works with their workflow, then who gives a [ __ ] right? Yeah. And and that was one of the things that you know working for Intersight and Alerna and you know we got to work with some of the big consulting firms and saw how they operated. Um and what was missing was a level of care about you know how the the entire team the entire like ecosystem would actually work together. When you work for a software company you know every problem can be solved by your software right for sure. If not, then like you're pretty incentivized to make it seem that way. Yeah. And and that was that limited scope was something that I found really frustrating because we'd work with somebody, you know, do a real process change and, you know, the change management was cut short because they didn't trust a software vendor to do it or the problem was this team didn't work well together or the data wasn't right or all of that stuff and that wasn't our job to fix. Um, and that that got really frustrating. So that was one of the first things when I started was for one I want to be somebody that these companies can actually trust to not screw them. And then the other part was need kind of a full scope of you know how to solve the problem no matter how what the solution looks like. Cool. No, that's that's awesome. I like that approach a lot. Yeah, it's fun. It's [laughter] important. Like I sleep better at night because I'm able to just do what I think is right. Yeah. Well, I mean, and it's so infuriating, you know, whether you're selling software or hardware into these companies because almost all of the time it's like, well, if we had this piece of data that I know you have and you own, but there's no way to access it or you don't want to give it to us or expose it to us or whatever, then it could make this process so much better. Or we could also do all of these other things on top of it if you just let us do that. But because you're in that traditional I'm just selling this widget software, I can't I can't push too much on that, right? That's [snorts] cool. Yeah. And you know, myself and you know, Josh and a lot of others, we would always do, I think, what was best for the customer, but that limited, you know, our careers sometimes because you're not doing what aligns best with your company and that kind of stuff was tough. So, like we've been really cautious about our incentives. So, we haven't taken any investors. We're we're bootstrapped. We're trying to we're trying to stay that way. Um, you know, we'll take money if we have to at some point to keep the lights on, but it's not looking like we'll need to, which is which is good right now. Um, knock on wood for that. Try to keep it that way. We'll try to keep it that way. And and and that just helps us have the freedom to do what we think is right. Yeah. And I think that's another [snorts] big struggle of a lot of startups, right? Is like, hey, this this isn't the most well-defined or well scoped project, but we need the money, so we're going to take it. and then it becomes this nightmare for everybody involved and yeah you just wasted a bunch of time and resources and all that. Tell me more about the the software kind of side the the actual stuff that y'all are working on that. Yeah, so the the tech side is really driven by a lot of the same goals as the consulting side of the company. So um for oil and gas specifically we're looking for gaps uh where either a vendor didn't want to build software in a certain space because it's not in their best interest. And so there's lots of those gaps that exist especially on the integration side. Um so there's lots of that and then um just from speaking to people sometimes uh we get ideas sourced from uh consulting projects that we work on and so we get ideas from that where people are facing real challenges and we have the technical skills to be able to fix that for them. Um so some of it comes from that for inspiration. Um yeah just speaking to people really like that's that's been driving the tech side of the business. And so what we started with um are uh focus on filling in those gaps like I said. So we have an Aries QC product and that was um based on speaking to so many different oil and gas companies that have quality problems in in their forecast in their production forecast. Um, and so we just created a little tool to go through and run QC checks. Um, that would make life a lot better for a lot of the techs that were running these queries every single month that would do that kind of thing. But, you know, it's always something that's not prioritized. So, they couldn't spend a lot of time building out more checks even though they know that they should exist. So, there's always a lot of manual steps involved in that. So, that was an opportunity where we saw we could just make life better for a lot of those people. So we built that. Um we put out another tool that's totally free that's called the upstream plan planning radar. Um and that was just to raise awareness of a lot of the software out there. Um if you speak to a lot of companies, they're not even aware that especially some of the smaller software companies, they don't they're not aware that those even exist or the fact that this software integrates with all the other tools that I'm already using. Um so we just wanted to make that available. So that one's out there and people are using and and yeah. Yeah. No, that that was pretty great. I know Michelle on my team used that pretty extensively. You we're evaluating some planning planning workflows, too. So, that was very helpful. And it's it's based on our own interactions with how people are actually using the software rather than what the landing page says. Theoretical conceptual design, right? Yeah. Because especially in planning, like if you look at the feature lists of, you know, a lot of the tools, they're basically the exact same, but who they're actually built for is totally clear for different audiences. Yeah. I can't I can't imagine oil and gas companies having any data quality issues whatso [laughter] Yeah. What a huge opportunity for you guys and everybody else out there, right? Because there's just so many when y'all are looking at it kind of from the production ops perspective, but there's all the way to accounting and legal and all that stuff, right? Like, you know, but I guess talk about you're talking kind of about the validator. It's what it's called, right? Pot. Yeah, that's that's the validator. And then I mean like that workflow [snorts] I mean can just shave you know minutes or hours of you know really not productive time especially if you want to iterate multiple times but you know you got to trust that exactly each time you run it but um and so how are you all kind of commercializing those? So one of the the mapping one is free y and then there's also a capex one. Yeah. Yeah. So, the validator was the the first one and then we had the radar and then the the newest one that we have coming out now um that we have a few customers already is the planner and so that's focused on one of y'all's LinkedIn post with the video. I was like that is slick. Yeah, that's what I said when he very sexy. [laughter] Describe that real quick because I want people to go check it out because it's it's impressive. Yeah. So, that's the planner. So, it's it's really for um we started with scheduling workflows. So um you know as you develop a field you need to figure out which wells to drill where and um all the time and considerations. So you know you have limited number of drilling rigs and frack crews that you want to run. Um so scheduling all that out so you can create a full drill schedule um that you can interact with. So that's one of the big distinguishing features of that product is that you can drag and drop live and see all of the other rules that you have set up. Um create a drill schedule for you around um all of the information that you actually know. Um so that was a big focus of this new tool new tool. And then um we're also building on planning workflows to that. So adding in uh capex so you can see as I make those uh right [clears throat] schedule changes, what does that do to my capex profile? So you can see monthly or quarterly or annually. How close am I to my capital budget? And then the the next piece that we're working on now is production focus. So you can bring in your production forecast and see what does that do to my production profile too. Yeah. So we're really excited about that. But Josh, my Excel sheet is super uh interactive as well, right? I I assure you it's not. No. [laughter] I mean like I mean you know uh full disclosure you know you know we at least been discussing some of these workflows with uh with these guys you know at GME and um you know I think everyone goes through their iterations I mean there's all different kinds of planning software I mean Microsoft planner you know Prima Vera and I mean in specific stuff for oil and gas and we've been going down the Excel rabbit hole and it's just like it's just not sustainable. Yeah I mean so [laughter] um it's how all of these billion dollar companies run on their operations side for the most part. Um, no, no, it is really cool stuff. I mean, so I think you know, kind of I think that's a pretty good segue maybe like into some of the data and technologies stuff. I mean, and again, we've kind of got a two-headed monster here with both you here, but I mean, as far as like getting into data technology or really maybe like heavy side, I want to get to you, but I was curious because you were on the math and applied applied math or statistics. Yeah, grad school is applied math. Okay. Like do you dabble in any of the coding or do you ever do R any of that kind of stuff? No, I mean I I did I did some terrible modeling in mat lab in in college and I did and I did learn I did some v VBA stuff when I had to as a tech and stuff but like yeah I don't I don't do any coding. Um yeah, so I'm gonna I'm gonna lean on on chat BPT like for anything I might do. Um but yeah, that's that's really Josh's Josh's space. like I'm really good at, you know, problem solving and and like that kind of stuff and designing systems and things like that, but you gotta know what they do. Building building a piece of of software. Yeah. No, you don't want me to do that. Yeah, fair enough. [laughter] I mean, that's that's exactly how I would qualify myself. Yeah. I mean, I think we all have certain levels of that though. I mean, I'm like to think I'm pretty good like in the data, you know, analytics space and, you know, I did a full stack web development boot camp, but I wouldn't hire me to build you a website. I mean, I I like to think that I'm a very good user of software. I can usually point out when, hey, like this is stupid and shouldn't work that way. But yeah, beyond that, yeah, and and that's one of the reasons our setup works pretty well is we're we're totally complimentary. So Matt's really great at figuring out which workflows we need to support and how to do it and then I can design an actual solution to that. So yeah, I think it works really well and and I'm I'm lazy as hell when it comes to software, too. Like I don't want to have to click more than like half a time to get to what I want. And and so between that and you know I think that's the other difference is like the difficulty with which Josh will add anything to the screen like it's it's really hard to get him to do that. Like it has to have justification and so like that's one of the things that we saw in the past is softwares just get really bloated right you know a customer asks for something you just say yes to get the deal like too much crap gets added to it without like that intention and yeah we're not going to do that. Nice. It definitely helps having both domain knowledge and then focus on like just generally understanding what good and bad stuff is from a design. Like I'm an ME as well and I would went to school to be an ME because I could the car that I had in high school you could tell that it was assembled by robots and then no one ever actually did any work on [clears throat] it. Yep. And it's like this could just be designed better so that this isn't nearly as difficult to change like the speakers was just a huge pain in the ass. And it's like they're two screws held in like why is this so complicated right? And so I've always been a big you know kind of proponent and just nerd about UIUX and yeah because especially when you start getting into analytics and data like you can view whatever you want in whatever type of plot and all that stuff but like the type of plot matters and where it is matters and how many clicks it takes you to get there matters and how many different ways you can get there matters. Like there's so much to it that I feel like so many people still underestimate especially in the oil and gas industry that like again best software in the world it can be best algorithm model whatever but if the user interface sucks and the end user isn't able to use it then doesn't matter that's something that was really cool um when we're creating the validator so it's for Aries users so it's for the people that have decided not to go to combo curve for something more like more modern and so like what Josh said is his design principle is He wants it to to be such that if they just mash the biggest button on the screen, it just works. Yeah. [laughter] Like there's one sales call that we had and like we taught the guy how to use the tool faster than we taught him how to share his screen. Like yeah, [laughter] that was that was a fun one. But I I think Joshells himself like he's he understands the workflows like he worked at an operator before too. So worked at Interside on the implementation side. So like he gets the business workflows but he's also like a wizard with the software stuff too, right? and taught himself as a kid. Um did probably illegal things. I don't know. [snorts] You probably don't want to talk about that. Everyone [laughter] those are kind of where I want to dive in though. I didn't know when you started, you know, your software journey like you know. Yeah. I was always kind of interested in software early on and so you know in high school I did a lot of just programming random things just for fun basically and you know everybody edited their MySpace page and that's how a lot of people learn HTML. Like as kids of the internet age or whatever we're calling that, millennials, whatever, like we we got to experiment with a lot of random [ __ ] that we would never have touched had it not been us growing up in that decade, right? Like exactly. I mean, why [laughter] I should have no re like half of us should have no idea what torancets are other [laughter] than all of the music down like you know the media and stuff like Napster and all that, but it's it's wild how that kind of just shakes out and kind of shapes you in the future. What was your first language that you learned? You remember? I think probably well HTML, but I I think like the first real language was PHP. Yeah. And it was just really popular at the time and that's what I started playing around with. Um just to as I started wanting to make dynamic websites, I got into that and then um and during my my engineering work terms too um there there was always an opportunity to write some VBA or something like that. So I I used to spend a lot of time spreadsheets, baby. Right. Well, yeah. I mean, you're handed the spreadsheet and and it's just staring at you and it's begging to be automated. So, [laughter] yeah. So, I spent a lot of time like that. I mean, I wrote a plugin for AutoCAD on one work term because uh we're doing a m material takeoff. So, you would like click on each line and measure it. And I was like, surely we [laughter] can automate this. Um so yeah I had a lot of experience programming through my work terms that way too just because it it happened naturally. Um and then so when I joined site I was working as a consultant there. Um but the same kind of thing happened and that you know we had um a lot of macros and things like that where it could just be a little bit smoother if we created a proper application for it. So that's where I started dipping my toes into it more seriously and um I eventually ended up moving from Houston to Calgary and I switched over to the development side of the organization and so I ran some uh development teams there and that's that's when I basically switched my career path fully into software. Yeah, that's cool. That's great. What uh let me ask you something. So, what kind of because I think what y'all do is a very like underappreciated or kind of overlooked approach to especially in the oil field. What is y'all's kind of uh pitch or just lessons learned so far with kind of you know the companies that you're working for and the stuff that you're building around you know hey this might not have a very clear like efficiency gain or revenue or I'm sorry profit increase or whatever but holistically it ends up making the entire value chain's life easier more efficient more accurate etc etc right like even something as simple as you [laughter] know showing the data quality issues in an Aries forecast like that potentially has millions if not billions of dollars of impact on some on a company right and so like something as simple as just hey this data is crappy so it needs to be revised is extremely valuable but it's one of those like chicken egg things right that you don't a lot of operators just get scared of because they don't know what they don't know until it it's too late or whatever but [snorts] yeah I I mean I we don't do projects you know, we're working with, you know, mostly publicly traded companies or very for-profit companies. And so if there's not a bottom line impact, like it's not going to happen, like it's not going to get funded. And and I think so we have to do is educate people on the impact of this stuff. So at even if there's zero you know change to decisions that are made or you know re I would say real business outcomes like you're going to increase your likelihood of people quitting um burnout you like they're like individuals effectiveness and especially right now a lot of executives are afraid of a recession so they're not hiring so teams are running super super lean well how do you get more out of your people like it's better processes and training right that's what you got to do um but yeah I I think the education and getting to the actual like value of the thing is something that a lot of people just stop short on. They don't really think about okay the knock-on effect of like the Aries data QC one like the easy one is yeah it'll save techs and reservoir engineers save them time but if you think about hey we're using Aries to evaluate potential acquisitions right and we know something about that other company's database that they don't like that's huge value and that happened like with one of the one of the customers and super cool which you know for a tool that's a hundred bucks a [laughter] Come on. Pretty good uh pretty good risk uh riskreward there. Yeah, cuz I mean even like you look at I mean we we just closed a deal for 825 million from a minute like even if you're off by a million you like that pays for itself magnitude you I mean yeah [snorts] for sure and and I'm actually I'm writing an article right now on you know how a healthy like work environment drives a healthy P&L um and getting into some of this stuff like good process happiness at work like a lot of that stuff goes together. Yeah. I mean, I' I've seen bad software, you know, at where we're at, you know, like that we've lost people like because they don't want to work. They don't want to use that software anymore. Like Yeah. Absolutely. Yeah. And I mean, that's one of the reasons on on the software side. So like how this plays into the decisions that we make on the software side, we we don't do concurrent licensing where you like share a copy across a bunch of people because on the implementation side, we've been sitting there with somebody, you know, on a Friday evening at 700 p.m. downtown and they have to get something out, right? But somebody left the tool open on their computer and so they can't get into it and so they're having to call whoever like Sally at her kids musical [snorts] to try to hey can you log out of whatever tool and you've seen that stuff and just like little things like that have a bigger impact than I think people think about. Yeah. Well, I think too there's like there's so many just holistically looking at it from a lot of my experience, there's so many like compounding things that like when people are analyzing it individually, they're like, "Oh, well, it's maybe not really that big of an impact." But it's like, well, the 20 minutes a day that it saves this person then makes the data quality better, so they have more or they have more time to make the data quality better, so now that impacts the person down the line that gets that data, which then impacts the forecast. There's just this a lot of it to me it's like a domino effect and a lot of times people only look at that first domino. They don't look at the whole thing. Right. Yeah. And and I mean our our main space is planning and like what what is planning? It's deciding how you're going to allocate your resources. And if like there was a study that we did years ago with KPMG where I think planners and people in planning said that they spend like 90% of their time screwing around with data or like manipulating stuff. And that's that's 90% of their time spent not analyzing things and not trying to make good decisions. And that stuff's harder to quantify, but it's real. Yeah. Oh, yeah. No, I mean that's uh I was doing some research on RPA and uh you know some of the the big consulting firms and stuff they were I was looking at it from an accounting perspective and it was I can't remember the stat which I'm sure I'll make completely make this up but I want to say it was around 70% of accountants time was still spent like manually entering stuff and as like a young tech person pulling down the Excel sheet doing things how how like because you have hordes of accountants at a lot of these operators and a lot of them are literally just taking field tickets and manually entering them in and it's like guess what the service company would love not to have to have field tickets. You would love not to have to have this giant staff of humans just monotonously typing in numbers. Why don't we solve this problem by using modern technology? And it's like there's so much of that but uh we could probably talk quite a bit on this. Yeah, I mean let's pivot. Yeah. So I can nerd out a little bit. I mean, so I mean, obviously some of this flows in like as they brought your design choices on being able to choose like technology behind your softwares that make it suck less, right? Yeah. Um, so I mean, it seems like you're definitely a bit of a polyglot, right? I mean, so you can kind of work in and out of different, you know, coding languages. I mean, what do you what do you say maybe typically attack a software problem with first, like C#.NET or something like And then if you need something that's faster, I mean, I know you're using some other stuff in um in the planning tool, you know, just just kind of curious where you like to start. What's in your what's in your stack? What's kind of your go-to for some of your, you know, what are some considerations of making those decisions? Front end, database, all that. Yeah, I I think a lot of the tech choices are driven based on the workflows. So, if we need things to go really fast, the the go-to for us is Rust just because uh we get really really high performance from that. And so anything that you're doing like a large calculation or analysis that might potentially take more than a second to run or something like that will always go to rust. And that's that's what we did for the planner and that's why the the calculation runs as fast as it does. Um but we use the best technology for that that workflow. So um on the UI I mean we can get the best experience just by using a lot of the typical web stack that's out there. So, uh, you know, we use TypeScript, um, for all of the the UI, TypeScript and React. I think that's like a pretty common stack for the front end now. Um, but I have experience working with uh, lots of languages like you said. So, you know, if I'm going to write a like a plug-in for PowerBI or something like that, then I'll I'll use C, which is like the the go-to for that, or same thing for Spotfire. Um, so it really just depends on what we're trying to do. um the performance benefits sometimes influence that and that's where we'd shift in one direction or the other. Yeah. with uh so with Rust um the only thing I can see as a kind of somewhat negative there is like you can't replicate yourself unfortunately or not yet at least right modern technology y um autog's [laughter] coming um soon you know is is like being able to onboard someone else or being able to find that that talent to to join if you need to scale at some point is that some you kind of figure it out when you get there kind of thing yeah I I mean I definitely thought about it a lot and so We have a developer working with us um who hasn't had any prior Rust experience um and he's able to pick it up pretty quick and he's he's a junior so um just cross trainining from Python onto Rust has actually been pretty smooth. Um, so I'm I'm excited about that because that kind of proves to me that you can probably pick up Rust at least for um, you know, some of the the simple tasks um, early on and then you kind of gradually get more and more into the complicated stuff, use it and solve problems with it and start to figure out. Yeah. So there there is a nice onboarding path there. I mean there it is a bit of a steep ramp for a lot of developers, but we uh, we work one-on-one and work through problems uh, a lot. So I think that helps a lot um especially in the early days when you're just first picking up a language like that. Um and the alternative is basically like I would have to use uh C++ or something like that and uh that's also very yeahve type safe right or y yeah it were C++ you have to manage the memory a little bit more and everything and yeah so Rust helps you a lot with memory management and and all that stuff. Um, so they're both statically typed, but C++ is just uh it's just a lot harder, I would say, for people to pick up versus if you have somebody that has exposure to uh Python or JavaScript or something like that, there there is a fairly natural path to picking up Rust, I would say. Um, and so our experience so far is proving that. Hopefully that's that's always going to be the case. Yeah. No, I mean that's a I was going to point that out. That's a it's interesting to hear because I feel like that's it's something a lot of people especially people kind of like myself or others in the industry that you know come from an engineering background but then get into data and like you know the logic of coding and trying to learn to code but then it's like okay well what do I learn first right do I do Python R C like where do I start and so being able to hear actual feedback from somebody that like hey Python Rust isn't that bad also helps helps, you know, from a dev perspective, right? That there's it's not future proof, right? Because that's that to me is like the biggest one of the big risks that I'm sure you probably run into with the operators, right? Is like we don't want a tool that's going to disappear in five years or, you know, we want open source because we want to be able to control it and not rely on somebody shutting us down or whatever. But that's uh that's cool. That's Yeah. And and Python's really popular in the in the uh industry, too. So there are ways you can um move just really uh high performance stuff into Rust and call it from Python and so you can have combination of Python and C# and Rust all in one project and that's totally okay. So just use the best language for the task you're trying to do basically most of the Python high performance libraries are really C++ that it's calling. Absolutely. And even like R calls a lot of forran and stuff like that. So um no definitely pretty interesting. So, and what about um I mean are your applications are they desktop cloud based? You know, how can they be deployed? So, we've done something that's probably not the norm for our our software so far. So, um we've started with desktop applications with a cloud component. So, it's kind of a hybrid of both. And the idea is that we try to get the best out of both worlds that way. So, um, you run your application locally on your computer, which gives you, uh, really fast performance, um, which we probably wouldn't be able to do, um, with a typical web app. Um, and you probably experienced this before. A lot of web apps just feel a little bit clunky for whatever reason. And, and sometimes it's because they can't, right, um, have access to APIs that you only have when you're running locally as a desktop app. So, uh, that's part of the reason that we've gone down that path. And then obviously having web apps is really convenient because you have the cloud component. So if you need to run a calculation across you know 20 servers then you can do that or if you want to have multi-user and you want to share projects and things like that it's it's easier for web apps. So um our thought there is we can start with uh all the desktop components and make sure that that works really well um and has a great UX and then um any of the cloud features that we want we can we can add it into that desktop app where it makes sense. So we can still share files even though it's a desktop app and a lot of applications work that way and we use Figma all the time and right you can you can save your files to the cloud and it and it works and so we can apply a lot of the same uh concepts uh for our our applications that way too. Um yeah and especially for the planner that's why we did it that way because you need that you know you interact with it. So you drag and drop a well on your on your GAN chart and then we don't want to have to go all the way to the cloud and run the calculation come back down because that adds a lot of latency into every single interaction. So um by running locally we can just do the calculation locally too. Um and we made it fast enough. So you know it can run on your laptop. Yeah. You know I have I ran on my MacBook Air this morning which is basically a potato. Okay. Yeah. Yeah. No, I was going to [laughter] I was just going to ask that question. um was just about like you know obviously people have varying you know degrees of laptops and and what about I mean I guess it is is it I guess because it's built on that web technology for the front end like it can run multi- um platform I mean Windows or Mac and yeah we've we've just started with Windows for everything I mean it can probably run on everything else too but we've started with Windows just because that is the operator operate on yeah yeah but I mean even if you wanted to be able to run calculations and and offload them. Um, you know, we're thinking about ways that we would bake that into the app, too. So, then you can switch if you want. So, if if your laptop is really really bad, yeah, for whatever reason, um, and can't run the calculation as fast as you'd like, then we'll have an option to be able to run it in the cloud. And then, so you kind of get to pick and choose like what what features make sense. I mean, you just you just gave a perfect pitch for what I used to sell, which is edge. Yeah, exactly. Like, it's literally the the the pitch of edge computing, right? is you bring the compute and the horsepower as close to the data as you can and then you kind of use the cloud as you know a integration abstraction layer or if you need to spin up you know a serverless load because there's some ridiculous heavy thing you can and then dump it back down. Exactly. And it' be totally fine you know you you get an edge compute and you put a box in the corner of your office and you just offload the calculation to that. So then you get to run on your laptop which maybe is low performance and you offload the calculation to edge um compute and then you get the uh latency benefits of being on the same network. So that's the power of having a hybrid app versus solely a web app. I mean I I sold uh basin modeling and reservoir modeling software for a while and the amount of I just never realized how much compute horsepower some of those like patrell and some of those have and it's like and even then like one of our differentiators at that time this was a while ago almost a decade now I think uh you know one of the differentiators was ours didn't have a CPU or GPU limit and patrol did right and so it's like you have these operators who literally have these data centers in their office, you know, that they're offloading these basin models to and they could only offload it to so many servers because it's constrained whereas the other one isn't. And it's like even then it would still take, you know, days to weeks to run these stupid models and you're like, I can't even comprehend [laughter] how much longer that would have taken if you were limited by the number of CPUs and stuff. So, it's like, yeah, they've got the a lot of them have these resources. That's really cool that you'll built your your infrastructure that way. Yeah. And and what's really interesting to me is how fast the average computer is now. Yeah. Um if you set up your calculation and everything in the right way, um you know, you can make full use of the CPU and GPU that are in regular commodity hardware and and things can go really fast if you uh if you're doing it the the right way. Yeah. Um and so that's what we're trying to unlock with with the planner is we'll make full use of the CPU and um you know, I have pretty strong background in in GPU programming, too. So, um, part of it's already running on the GPU and we'll probably take more advantage of that in the future so that we can take fast or take advantage of fast GPU compute. And I think some of this ties in because you've got an open source project too, right? That is kind of born out of stuff you're doing for the planner, but that is actually getting some pretty good traction. And that is that based around the uh, WebGL stuff. Is that right? Or Yeah. So, there there are a bunch of open source projects I work on um, that are related to graphics. So um some of them are focused on WebGL and there's an upcoming standard well actually it's just being stabilized in Chrome now um called WebGPU which is kind of an advanced version of of WebGL. So um you know there you've probably heard of DirectX and Vulcan and and things like that. So WebGPU is trying to take advantage of the next generation of uh GPU APIs and um so I've been really involved in web GPU. Okay. um since it was a first proposed and we're basically working on the implementation of that and so through that that's where I picked up a lot of my GPU experience and and that kind of thing. Nice. So like concept of that just to allow allow calculations on the GPU from the browser is that the simple way to boil it down for Yeah, exactly. So um graphics and compute. So you can already do um graphics through WebGL in the browser and you might have seen some demos like that where I actually use it in plotly. Um plotly does it because they have the two WebGL call and then you can make you can plot like millions of points. Exactly. You just get a canvas in in your website and you can render these really advanced 3D visualizations directly in the browser. And so that that's already possible with WebGL. WebGU web GPU is kind of the next version of that. So um so plotly for example would probably uh upgrade internally and then you just get way faster visualizations and then on the compute side it means yeah you can run um compute directly in the browser. So there's lots of different use cases for that but one of them would be um you can do really fast matrix multiplication right so and that's really useful for all of the ML use cases that that are out there now. Um, so it makes a lot of that possible. So you can run really complicated ML models directly in the browser. That's like Yeah. Yeah. The browser's an interesting space, right? I mean like because even like I mean you know duct DB's become a big thing on the like OLAP data engine sites like SQL life or that and you can run that with web assembly in the browser and then was it Anaconda put out the whole like where you can run Python in in the browser. I mean it's it's going to be interesting where people kind of decide to make these design decisions like server side or in the browser. Yeah, the browser is basically a full operating system at this point. So, I was just trying to figure out how do we create an operating system that can run on all the different uh real operating systems out there. It's also the reason Node.js was spun off that, right? It was just the V8 engine that exactly Google had in Chrome. They just ripped it out and said, "Let's make a server side, you know, framework out of it." Yeah. So, it's the intersection of all these technologies coming together. So, you have web assembly and web GPU and everything um all kind of happening at once. But I it Yeah, it's really interesting how that whole ecosystem is gonna evolve over the next year, probably even few months. I know it's it's so fast these days. It's so crazy. I feel just saying that out loud makes me feel old and I'm like [laughter] back in my day, you know, but it's like [ __ ] every week it feels like there's a new thing, you know? Yeah, it it was interesting whenever, you know, we started with with products and and Josh would tell, you know, some of the folks say one generation above us on on the software game that, you know, we're not building a cloud app. They're like, "What? What? What are you talking about? What are you? You're going backwards." Yeah. And building an execut I didn't understand fully why. And I, you know, trust Josh and and everything. And then, you know, as it comes together, I'm like, "Okay, I get it. Your uh your P&L appreciates it greatly, too." Well, that's the beautiful thing about anything we can, you know, any way we can avoid giving Jeff Bezos more money. It's the the stat that I I know from my edge computing days was it's something like 80% of startups GNA is spent on cloud. Yeah. And it's like because I did a I found this case study that Dropbox did and they cut their cost by like 74% by doing like a hybrid uh not true edge but you know kind of regional edge type deploy or colloccated deployments and they cut their cloud cost by 70 something%. And it's like that's insane, you know? And then, you know, piggybacking off of that, right? It's like then of course makes that makes your life a hell of a lot easier to go sell people because you don't have to have this GNA just overhead all the time just to host the stupid thing, right? How many startups failed because of that? Yeah. Well, I mean, right, like because a lot of the a lot of the pitches, right, are well, once we hit critical mass or scale, then those costs, you know, go down exponentially or whatever. And it's it's like well that's a big if. Yeah. [laughter] In the space. No. And I don't want to go too far down that rabbit hole, but I think you also made the case like there's also people that were able to start off and you know build companies because they didn't have to have put server in their closet. Is it's a trade-off for sure. I mean one of the one of the parts about it is we do like having the lower cloud cost because we can pass those savings on to all our customers. So that's a big benefit for us especially being a smaller company. Um the other part I didn't mention earlier. So I talked about the latency of you know the full you know the hop to the cloud and back. Um but the other aspect of that that is just moving data and and having control of your own data. So that's also really important to me. I know how complicated it gets with some of the the new software that's coming out now and just trying to pull your own data back down from the cloud is is it can be really painful if it depends on the vendor entirely. You know, I'm not saying all vendors do this, but a lot of them will lock it in and you pay like an extra license fee to have access to the cloud or you you're charged based on bandwidth usage or um they don't have the APIs you need to automate things that you'd like to. Whereas that's kind of like a step backwards from um the totally raw database access that that was kind of the norm for a lot of the on-rem apps. Um so I think that's an important consideration too. And so that's something we're really aware of. And so even if we do offload compute to the cloud, we want to make sure that you can still um grab all that data really easily and you don't have to wait. Um and you know a lot of the internet connections that operators, it's not the best. Mhm. Well, and even then, right, like we were talking about this on a prior episode, it's like, is calling an API that returns, you know, a thousand rows at a time for millions of rows of data every single day across a bunch of different assets more efficient than just duplicating the database. And Exactly. Yeah. Yeah. There's that. I mean I mean sometimes like man I wish I just go back to FTP and just [laughter] and the the data ownership side that goes back to what you said earlier about you know a product that might not be around in 5 years. So whenever we get into projects around data management or or software and stuff like that like my advice is to is for companies to really own their data and own the integrations so that it doesn't matter like if a software sells to another company and they kill development. Not like that would ever happen, right? [laughter] No. Or get bought by a major service. Yeah. Like I if if I'm running an ENTP, I want to not care about all that. I want to say, you know, I can use the best tool on the market at that time and, you know, if a new scheduling tool comes out that's better than the one I have, I want to be able to swap it out and, you know, get the data. I mean, that's that's the standard that, you know, if you look at the tools that like software companies use and tech companies use, like HubSpot and Zapier and like all of these tools, it's integration and all that stuff is instant. Yeah, I know. I [laughter] I fought for so long against using Zapier just out of principle. I'm like, I'm just paying someone for this API that I could write, but then I'd have to manage and like all that stuff. And it's like, you get to that point where you're like, it's just worth it's not worth my time. Yeah. I just click a few buttons and it's done automatically, right? But it also, I mean, from y'all's perspective, it's really smart as well, just from a risk and security perspective, right? Like that was one of the easy things for me when I was selling edge is it's like, no, it's yours. It's on your network. You control it. It's behind your firewall. You get to control everything, right? Whereas, you know, if I'm trying to sell a cloud solution, that's a much harder conversation with an operator because every single IT department is going to be very leerary of data exposure and risk and sock 2 and all the other stuff that comes with all of that. But yeah, so one thing I want to get in and in, you know, kind of a stack uh a stack question, but it's more about so I know you guys are really big on kind of the work life balance. You know, everyone's distributed and remote, whatever. um is kind of like your communication stack. I mean, yeah, thought about, you know, I've looked into it a good bit. I mean, do you guys ever read like the GitLab remote work manifesto or any of that stuff or if you have, it's great. Or if you haven't, you know, definitely worth checking out. I'm sure you guys are a thing or following some of those principles, too. But just curious how how you guys communicate and what you use to facilitate that. Yeah. What's in your ops comm's stack? Not much. Yeah. Um, so I read like Base Camps stuff. I really like their their kind of model. So our our communication is almost completely asynchronous. Um so Rome research is the tool that we actually use. Um so it's it's a note-taking app on the surface, but it's really um it's got a lot of depth to it for connecting ideas and connecting um you know topics and and things like that. So you can link in any direction that you want. And so it's really like built to be like a second brain. And that's what we kind of see it as. It's our brain, but the majority of our communication happens there also. So there's a daily notes tab or a daily notes page that pops up. Each of us say, you know, here's what we're working on that day. Um, so we all see it. And so like we're not going through and reading or telling each other what we're working on, but we're all just like aware, right, because we just see it in the periphery where we're working. Standup call. It's Yeah. Five minute stand up. Yeah. Yeah. And so um, so that's the main thing. And and what's cool about it is, you know, I can I can tag, hey, inbox Josh, ask him some question about a company or a feature or whatever. Um, and if I tag it the right way, then next time that feature comes up or that company comes up, you know, we can see all of those other notes and all that history, like it it it's it's a note-taking app that works the way your brain kind of would, right? There's like a discovery kind of searchability. Yeah, it's really really cool. And you can add structure to it and we've done a lot of that kind of stuff. Um, but that's the main one. So, we don't re we don't do Slack or Teams. We don't do instant messages. If something is urgent, we do like text or call. Um, and then we don't really email each other like hardly ever. Like if there's papers to sign or like other stuff, then we'll send things around. Um, but yeah, I mean that's that's basically it. Interesting. Heard that. It's Rome R. Yeah, Rome research. R O M Research. Um it's not it's not very popular for like companies to use. Notion is more I was gonna ask if it was similar to notion aspect that it's like it's a box with a lot of tools in it and you can configure that box however you like within the framework of the tools. Yeah. I I'd say that it's more of a box with like the materials that you could make tools out of. [laughter] So it's like less structure, less like it's very open-ended. Like people use it for journaling. Um, other people use it for like actual research, like doctoral studies and all that stuff. Um, the way that we use it, I think, is pretty unique. I've connected with a a handful of other founders um, around the world that use it for their company. Um, and it's really really powerful. Like notion is a little bit more structured. Um, what Rome is really good at is rem removing any friction from you typing stuff because that was always my problem with notes is like I open up one note and then I have to think about okay what is the topic and then what folder should that sit in where does it sit in the hierarchy and oh it actually fits between these three things and then oh [ __ ] I forgot what I was going to write in the first place like before I even get there but Rome you just start going and then add structure tag stuff later. That's awesome. That's pretty cool. That's cool. No, I mean we use Notion quite a bit and ClickUp even is starting to [clears throat] add notion like features and stuff like that, but that's uh that's a new one. And what Microsoft has to be adding their notion competitor are they doing? Yeah. Um I mean, you know, they're always going to be a fast follower with that stuff, but with their GBT, you know, yeah, I mean, notion co-pilot type stuff. It'll be interesting. I think I think Notion was one of the first ones right after the GPT3 release that I saw like actively integrating some kind of AI into their notion or into their software directly. And it's okay. I mean, it's I'm I can't wait until the GPT4 API goes public, which please make that public as soon as you possibly can. um and also let us run it on our own servers so that uh we can have a lot more fun with it. But um it's I mean those productivity tools are so important in my mind because you know even someone like we're using notion right notion integrates with a handful of things but then it also doesn't integrate with a handful of things right and so it's like having some of that flexibility on the back end where I don't have to go stand up a stupid zap just to get whatever from somewhere else is so nice. Yeah. And a lot of the productivity stuff, it's there's a lot of masturbation like snuck into it. Like in to-doist, like you're incentivized to create as many checkboxes, so you can check off as many check boxes as you can and stuff. So that you're busy. Yeah. And and Rome, like there's there's no notifications. Like you can make a checkbox if you want, but unless you search for it or have a query, you won't find it again. And so like that freedom is is pretty cool. And because I I think we really wanted to build stuff around preserving focus. Yeah. for the individuals and you know the most value that we can create is by us creating things. So him creating software, me creating ideas and and and other things like that. Um yeah, Rome research it's pretty wild like the founder is like out there. So he he went and lived in a van for like three years to come up with the idea and like create the first version. Um I talked to Yeah, there's so we back it up all the time just in case because he seems like the kind of guy that could just be like done. Yeah. Um, nuke this. I'm No, I've talked to him and he's got some really fascinating like ideas because the way that it works to connect like an individual's brain and thoughts and ideas. He wants to make that happen for like everybody. So like connected thought across like everybody on Earth and he's got some wild ideas. So it's not in his mind it's not a productivity tool or note-taking app. He doesn't really care about that stuff and they're funded for 10 years. Yeah. So, he doesn't care. Um, future's going to be crazy. Tana or Tana is another one that popped up that looks really, really good. And they're more, you know, traditionally structured. I think they're like Dutch or something. Um, I haven't tried it, but I've heard really good things and and it's a little less uh wild west than Rome is. Yeah. What's the So, I like everything except when you said you don't have Slack and that made me like cringe a little bit. I mean, they don't have the upstream people Slack, but Yeah. Well, I know. Well, that's why like I assumed you use Slack because of the channel, but what So, how do y'all do like you know DMs or groups or just that like either single person coms or small you know department Yeah. comms. I mean, we don't really like like we have two like we set out our communication principles where if it's urgent like text or call um and we use WhatsApp usually for that so we can you know have a group call and stuff like that if we have to um and then the the messages that we send each other in Rome like usually we're all working like during the day and so so you can like DM in Rome. Not really like it's public. Okay. Like within Rome just like you're tagging people basically. [clears throat] Gotcha. Yeah. And and so that's usually where a lot of our like back and forth messages are. Um but like I said, there's no notifications. So like if I'm focused on something, I won't get a ping that ping ping. Yeah. It's not like HubSpot telling you every time someone hits the [ __ ] website. [laughter] Yeah. Cuz like mo most of our stuff like doesn't need to be answered immediately. If it does, call because that'll break through everything. And then we have two meetings a week. Um there's a book called Traction. Um, so we kind of took some of those ideas and splended it with our own stuff for meetings. Um, so we have one on Mondays that we call business time. Um, that's where we look back at last week like briefly talk about this week and then we get into like issues. Um, and we'll we're very like individualistic. Um, so we focus on like, hey, what would each of us want to do differently this week as opposed to last week? Like what habits like work habits are we trying to like instill and get better at? and how do we spend our time and be real intentional about that stuff. So, that's more the the topic of that Monday meeting. Um, and collaboration on, you know, projects that we're working on together and stuff like that. And then Friday is a headlines meeting. So, throughout the week, if something really good or really bad happens, we tag it and note it in Rome with and we tag it as headlines and then it pops up and in in there. Um, and we just reflect on the week and sometimes if we have time, if there's not like something really big happening, which lately there's been a lot of really big happening all the time, like in a good way. Um, we'll talk about like ideas and random stuff for the future and things like that. That's cool. Yeah, those are the only two meetings that we have regularly. I think you and Chris might have like one other one on the on the dev side, but yeah, like like I mentioned earlier, sometimes it it is useful to have the one-on- ones and just work through, you know, we do a lot of pair programming. Um, if if some of us get stuck on a problem that we just have one-off meetings for those sometimes, but we don't have a lot of recurring meetings on our calendar or anything like that. We're trying to keep as much focus time as possible. And you know, part of that is also that we're spread across time zones. So that's not even really feasible to do even if we wanted to. Yeah. So we're pretty careful about preserving time. And so Rome, like Matt was saying, is really trying to get the closest we can to a mindm for the entire team as possible. And so whether that's Rome or another tool, we don't really care about exactly which tool that is, but just like the idea of having one place where we can have all of our ops, that's really important to us. No, for sure. I mean, that's I feel like that's what a lot of the individual tools out there are trying to do. Yeah. Right. Like Teams or SL, you know, Slack and Teams is a great example, right? It's like okay well this is good for like this very specific thing but there's also this other thing like document like notion provides right where you can tag and embed and create tables and stuff like that teams doesn't like and so now okay Microsoft comes out with a new product for that but now there's got to be some kind of inte it's a lot harder instead of just putting it all in one spot and letting it organically kind of tag or doing it the way that it makes it's interesting. what do you use for kind of dev uh management and um so we're on GitHub so that's where we put all of our projects and everything and um like to Matt's point earlier about software teams being able to swap software out easily you know if GitHub disappeared we're on the bigger problem well but you yeah since we're using git that's the underlying protocol that we switch to GitLab or you know there's all kinds of services out there so um we're using GitHub for that um which works really well since we have a lot of the open source projects there. Anyway, um for the actual devstack, you know, we're we're using VS Code, we're using C-Pilot, which is yeah, really cool. See, see, today is VS Code day for those who celebrate. [laughter] If you wanted to sponsor the podcast, we are we are welcoming. But yeah, no, I mean yeah. Yeah, this I mean talk a little bit a IML and I know we probably wrapping up here in a few minutes, but I mean like yeah, the co copilot has been pretty awesome and I've been interested to hear because I use it primarily with Python, but like I've talked to someone who's use Go, you're using Rust. I mean, it seems like it's working pretty well across most it works well for basically anything. I mean, as long as it has some data set [clears throat] with the programming language that you're working in that it can kind of figure things out. And I really like it for the any kind of uh more tedious tasks, especially just autofilling things in. Um you know, you have a table and you're you're just trying to you don't want to write like a full macro or anything to like copy the lines and insert data, but it can autofill information really well. Like I used it the other day for I just wanted a sample data set for wells and like you can literally type oil well and it it has that concept and it can start filling things out for you and it can create like a full JSON document of like the properties that you want for oil well with sample data which is really really cool. I mean I think that's one of the talks because this is a whole like I think it started at noon today central but like I think one of the talks was about generating like sample data sets with copilot. Yeah. In in VS Code like yeah it works really well for those kinds of use cases. So that that's what I really like it for. And yeah, just it's like a more powerful autocomplete basically. Yeah. No, that's I mean that's something I've been waiting like 15 years for ever since I started coding. Yeah. I mean literally you can write comments and it gives you like pretty much functional code. Yeah. Yeah. Yeah. No, I mean that's one of the main things I use GPT for these days and started messing around with Bard with some of that stuff, but only because my hypothesis is being a Google shop with Bard, maybe it'll understand collab better [clears throat] and so I can just dump it in there and then run it and collab easier. But uh what else are I mean are you guys are you using GPT or any of the new fun ML tools at all either personally or in Not right now. I mean, we've I think everybody's played with chat GBT and I mean, obviously, it's pretty mind-blowing what it what it can do. Um, and I think for the actual day-to-day, it's just co-pilot for us, but we're definitely we have an eye on it. I think there's a lot of interesting use cases that we can start applying ML to in in in this space. So yeah, it'll be really interesting over the next few months because like we mentioned earlier, this technology is just like taking off so fast and it's it's honestly hard to keep up with it dayto-day. Yeah. Well, that's like I was at a AI meetup this weekend and they were one of the guys is working on vectorizing some PDF data basically and uh you know he's like well there's six different vector databases. They're all brand new. I think this one or I'm using this one, but that's only because that currently is the most popular one and it's like I don't know if it's gonna be here in six weeks or eight months or like you know so it's it's still it's so early. It's such kind of wild west with a lot of that stuff. It's it's fascinating. I I like GPT just because it'll document the code properly every time, you know, like just the little things, right? Like that's better than a junior dev who isn't going to properly document it all day long. Yeah. Right. I think I think using something like that in the planning space is pretty interesting too because I think since Alexa came out, people have been saying they wanted Alexa but for oil and gas. And so being able to say like, hey, what's this month production look like? Like all of that kind of stuff. Like I've heard multiple executives say for a while. Yeah. Yeah. Yeah. there's that that need has been or the desire has been around for a while but you know we're going to think about how much of that is desire an actual need like some of that stuff is just ego stroking like cool but some of it could we were talking to you know another guest and he was like he's like even if I'm using it I'm not going to say that I'm using it because I don't want to like cheapen no I mean it's just like I mean like just use it to enhance the product but right as soon as you start branding are we using AI ML just cheapens what the actual work the software potentially gives you know that bad taste of overhyped promises and underdelivered results. Yeah. Blockchain [laughter] and and yeah, I think there's a lot of people that use it for like there's a lot of companies that started using chat GPD for like content and like blog posts and social media and all that and some people seem to like do that pretty well, but a lot it's very obvious. Yeah. Like like it's very very obvious. Yeah. No, there's definitely a a mastery to be had in just prompting, right? Like I thought it was crazy when it when I first started seeing people talking about jobs for as prompt engineers for GPT and stuff, but it's absolutely in my experience playing around with it, the more you mess with it, the better you start getting at like even the just like knowing how to Google properly. Absolutely. Like do you even know what a quote does or the word and like it actually matters in us Googling versus our parents, right? Yeah, it's going to be unfortunately the same thing with [laughter] AI and GBT long and like I do a lot of writing and stuff and and like I take a very like creative approach to it. Like some people are very like methodical to writing and outline and like topic by topic and so I found that it helps me like be more comprehensive and like outlining and so like I'll have an idea or a message that I'm trying to convey and I'll ask it for a like give me some counterpoints, tell me why this isn't true and then I can think about those and then write something based on the idea. So, I found it useful for some of that, but yeah, it sense of humor is terrible. Like, I tried to like use it to say something funny and it was just bad. Yeah, it's it was very like corporate. It definitely lacks like the human nuances in a lot of scenarios. That reminds me like was it South Park or the Simpsons one had like a it was like a German robot like a joke bot whatever and it was just terrible and I was like [laughter] well that's like Sarah Sarah was telling me last night she's like have you seen the the Snapchat AI everyone's talking about and she's like I don't understand she was like we've been doing this since smarter child on AIM and she's like I don't understand why people are so excited. That's probably [laughter] a lot better than smarter child was on a but again to agree you know I think one theme we're seeing a lot through these podcast too is like everything old is new again too like yeah absolutely well and I agree with you wholeheartedly right like at some point it will probably get to the point where it's kind of like the deep fake stuff almost indistinguishable with blogs and content versus people but right now it's way better at like hey I've got this idea give me some talking points points or an outline or here's my I here's my writing make it not sound like I'm an idiot or proof it debug I mean you use it for code debugging too like just help me make this better than it currently is by myself you know yeah because to me it's kind of replaced the gap that Google left because it got overrun with ads and SEO and just the quality of your search results just went bad so seems like chat GPD's like filled that need pretty as long as as [laughter] long I've started having to do this putting like a prompt at the end of it saying like if you're unsure of the answer tell me you are unsure because it will completely feed you [ __ ] that you didn't double check or go you know look it up you would be like oh sure and then like I was trying to look up some company data I was like let me see if I can like scrape some stuff with this just to see like what it would do and it was spitting out you know I was like give me the street address city state zip code URL LinkedIn stock ticker number of employee, all this stuff. I was like, I just want to see what it returns. It started and gave it to me in a tabular format exactly how I wanted and it filled everything out and I go to LinkedIn and start looking and I was like, none of this is even close to right. Like I was like, how did it even come up with this? You know, it just spits it out. And so that's my AI warning for the show. It does apologize if you correct it though. Yeah. [laughter] You want to jump into the last question and then speed around? Yeah. Um, yeah. So, what's, you know, I guess for both of you, what's one piece of advice you'd give to people looking to get into energy tech or the current energy tech community? I think there are still so many gaps out there um where if you deeply understand people's workflows and you work directly with them to understand exactly what they're dealing with, then um there are lots of opportunities for software improvements or I mean on to Matt's side, like all of the process improvements that are out there. um there's still plenty plenty of opportunities. So, um I would say try to find somebody that you can work closely with. Maybe it's a company or just some some engineers at a company somewhere. Um work closely with them and see their day-to-day and there are all kinds of opportunities still. What about you? Yeah, he he stole my answer. No, but I I'll build on it though because I think a lot of companies they do a good job of finding like the initial idea and finding a problem that they need to solve and then solving it pretty well, but then get really distracted after that. And so like keeping that question of like what is the problem that you're trying to solve top of mind all the way through? Um I think that baseline, that foundation of like actually doing something valuable, I think can't go wrong. Yeah. But I think to your point too, like I feel like because just society, right, we see a lot of these tech startups and unicorns and all this stuff, everyone coming out of high school or college wants to go create the next big thing. And it's like there are so many more companies that were created because someone went and worked in an industry at a company whatever for 6 months to 20 years and saw through that experience that there was this problem and this gap and then they went and fixed it, right? And so it's like what I'm trying to say is you don't have to go out and become an entrepreneur right out of the gate. like go get some experience and learn and see where these gaps are because it's a hell of a lot easier to go fill a void that you know exists and you're directly experienced with and understand the stakeholders and the processes and the software and the workflows and all that instead of just coming up with a random thing that you think might help. Yeah. Yeah. Go get a customer support job. Yeah. I mean that's where you'll find lots of business ideas in there and problems that whatever company you're working for is not solving. Yeah. Y and people in the industry have been really great and everybody's willing to share. Yeah, these are my big pain points and I I would really like this aspect of my job solved. Um so everybody's been really great. So just ask the question, phone somebody and or 20 people and and see if they have common problems and and build that exact solution. Um like Matt said, be focused on that exact problem and don't get distracted by all the other shiny things out there. Yeah. I mean, one of the reasons we're building the planner is because a lot of people asked us to. Yeah. Yeah. [laughter] So, it's pretty clear sign. Yeah. Yeah. We would show we show them the validator and they say, "That's cool, but can you build me a a planning tool instead?" [laughter] Yeah. But now you got to get down to, you know, at the customer level and understand what what they need and what they want and and I think, you know, kind of similar but different, you know, you're not valuable because you know, Python or right or R or whatever, you know, like you're valuable because you understand the business problem and then you can use one of those tools to solve it. You know, like you see a lot of people, oh, I did a this boot camp or whatever and now I know Python, so you know, I should be hireable. It's like no, right? you know, I'm valuable because I can go talk to someone in accounting and tie it to someone here, you understand why why that matters and then, you know, build a proper solution around it. I mean, most people again, they don't care, you know, most people that will ever use Josh's product will never care that he wrote it in Rust and whatever things behind the scenes. It just it needs to work. It needs to work fast. And um so I mean really you're solving their problems and they don't care how you do it. Yeah. And and that's one of the I think misconceptions is that when we're talking to somebody and they ask, you know, could you do this? and we're like, "Yeah, if you want to." Um, I think they sometimes that comes off as, you know, we don't have an opinion for how it should go and we're just kind of open book, but like what's actually happening is like we have a very strong vision for the end point, but it's up to individuals convince you otherwise. Well, that but also each person's like path to get there is going to be a little different and how individuals like adopt technology, how comfortable they get with things, like that kind of stuff. So I think we've got a vision for the endpoint, but you know the way that different people can make it there, that's that's how to manage the change. Yeah. No, I mean it makes so much more sense. I think just generally speaking, especially for the kind of generalist management people that aren't in the dev day-to-day, the languages, the databases, those are just tools in a toolbox, right? Like the dev themselves is the one that is going to be doing these things. So having domain knowledge experience etc is so much more valuable than if you know a dozen different languages but you have no you know production engineering experience right like it's it's something that I feel like goes overlooked or like you look at the some of the job applications and stuff and all the requirements for must know all these languages and stuff it's like is this person actually going to use all of these languages because the description doesn't even make sense in that context right yeah I think it's the intersection of you have software knowledge and the domain knowledge which I think is so important and so if if you have a background as a developer then it it's worth trying to understand the domain really deeply and and same way if if you understand the domain really deeply and you want to get more into the software side then I think that's a really powerful combination it's like a superpower for a developer to actually understand deeply the workflows that they're trying to solve especially in a science and engineering world where people also want to debate the differences in a physicsbased model and a statistical based model Right. And like I also think just generally speaking as engineers we're very logically driven as it is. So that's why to me coding is relatively easy understanding code is relatively easy. All the detail and stuff that I get hung up on can be solved with things like copilot and whatever else comes in the future. But yeah just having you know for all the engineers out there that are kind of like wondering about it. It's new and it's scary, but it's very much so in like it will fit in your brain very easily because it's just logic. That's all it ends up being at the end of the day. A lot of a lot of it at least. Speed round. Yeah, let's jump in. Go first. All right. Just short answers. We're just going to bounce a bunch of random questions off of y'all and kind of go from there. Uh, what's your favorite database? Oh, I have a tie between SQLite and Postgress. That seems to be a common theme. I'm gonna say one with unique identifiers. Yeah, [laughter] very very not just uh long form text fields. [laughter] Those are my favorite. Yeah, just join on well names. It never never sees API Bobby. [snorts] Um yeah. Yeah. So, what's your favorite, you know, managed service or something that you guys I mean, I guess kind of hit on it with Rome, but yeah, I I think for our actual ops, it's Rome for sure. Um, I mean, even though we talked about the benefits of running locally, we do have AWS servers. So, yeah, a lot of their services are are great. I mean, the UX of AWS isn't the best, but the services themselves work pretty reliably. So, you know, we have things like O running on AWS. So, um, all of their managed services are generally pretty good if you know how to use them. Yeah, I'm still I can't wait to see how they solve all any of the cloud providers solve the UX problem because it's Yeah, like it's not an easy thing to bite off when you think about all of the different services and tools that they Well, but even that like when you talk about UX, I mean, so I think that's really where you get separate UI and UX. Like a lot of the UIs are terrible, but then you have to remember they also have a whole like UX like developer experience like CLIs and SDKs and all stuff that maybe are actually pretty good. Yeah. But yeah. Um but yeah, I mean again AWS is just overwhelming. You lo all of them are like oh an engineer clearly had his hands in this but at the same time it's like how else could you do it too? Right. You can see how it happened. And I mean they started with like a handful of services initially. So you click the menu and you just see those services. But now you open it up and your entire screen is filled with all the service names and you don't know what any of them mean. So yeah, I think that part could probably be better. Yeah, I'm going to throw a curveball. What is something that you see as having the greatest opportunity for improvement? And this can be tech side for you, kind of more operations management side for for Matt. [snorts] So to give you an example, my answer was uh around APIs and data structure or no, I'm sorry, it was around consortiums and how I think it's complete [ __ ] that we've got all these oil and gas consortiums that are tens or hundreds of thousands of dollars and they're all moving towards the same thing except they never make any progress. They never do anything. and then you're just by default squeezing out any potential new companies or small companies by having some kind of ridiculous price tag like that, right? Like I'm personally of the belief that it doesn't really matter what the actual structure is as long as you define it and you document it and you stick to the structure whether it's JSON, XML, whatever, like that's going to change because those tools are going to evolve over time. But just have coming out and having like a standard uh you know frack naming convention like why doesn't anybody publish that or just have that right like that it's such an easy thing but then everybody wants to have their own proprietary one or whatever you know anyway that's kind of just context I I think mine is actually the same I don't know if that's allowed but no absolutely yeah I this is something I I've brought up a bunch of times you know there's no reason that we have to have our own convent vention for reserve category. Um I mean you can define that once and you just share it across companies but everybody wants to code it slightly different way. Um and it just adds a little bit of friction every time you have to map those even though it's representing the exact same thing and you can apply that to so many different concepts. It's reserved categories. It's area field. Um what what are you naming all those things? I mean naming is a whole another it doesn't even matter to to anybody. So, it's not because they've intentionally named it slightly different than another company, but it it just adds so much pain when you're just trying to map the same information [snorts] over and over and over. Yeah. I mean, that's why you brought not to you brought that up, but I think was it in the Slack channel about the AFS like like why why do all the categories not match across companies? I mean, there's only so many ways you can possibly spend capital in oil and gas. That's the like so I got my masters in energy business a couple years or five six years ago now and when I took uh finance and accounting and I realized that it's just like wide open you know as an engineer I'm like very structured and like you have physics and you have all these rules and stuff that you have to abide by. My mom is an accountant, so she's very like anal about the way things get done. And so I always thought for the longest time that accounting had like these very strict rules and like generally speaking, they kind of do, but then it's like, well, you can pick which type of accounting you want to do out of the gate. And then based off of that, you can do whatever the hell you want essentially under it within this, you know, framework. And it's like, you mean to tell me there's no standardization in this one industry across just accounting principles? It's like, well, it's these two methods and then it's just the whatever you want to do within that. It's like that's crazy. I think mine mine's related and and but it's around like buying and selling habits for software in the oil and gas industry. I think buying habits are super relationshipbased and you know which you know person brought you pies most recently and like a lot of that kind of stuff. And I think the the vendor side of that problem is the transparency into you know what does your tool actually do like putting demos on the website pricing like being more open about all of that stuff. I think I think the entire industry and the market would benefit um but that takes people too. Yeah, but [laughter] that that takes that takes [clears throat] people, you know, taking a more collaborate collaborative approach to things and more open and stopping like trying to kill each other and eat each other's lunch and No, they're all doing it. Proprietary stuff, Matt. Everybody's doing something completely unique and different. Yeah. Like they got people wondering if someone's getting over on them and like wait basically the same company as them and they're getting a 50% discount from us and people talk you know you know behind the scenes like you know whether it's talking to other vendors talking to each other or operators like your pricing is never as secure or as secretive as you think it is in the oil field. Absolutely. Definitely not. And and I mean that's the way that we kind of approach it is like when when we're doing sales it's very much here's what we have. Let's understand your problem. see if there's a match. If if so, cool. If not, that's cool, too. Yeah. Yeah. Thank you guys for coming out. Where uh where can people find you? I also want to give you a second to talk about your book or tell people about it where they can find it as well. Sure. Yeah. So, you can find us on the internet all over the place like LinkedIn, um Twitter, and then pod2.co is our main site.co. Um yep.co. And uh so pod number2.co. co. Um, we've got a YouTube channel. So, our podcasts and stuff are going up there that we recently started. Um, yeah, I wrote a book called Upstream Integrated Upstream Planning, the essentials. Very wellreed on LinkedIn. Lots of No, I've I've heard really good things. Like I got a message from somebody yesterday that I don't know, it just brightened my day. It was It was really cool. Somebody I'd never met and they were like, "Hey, read this. It struck a chord with me. It was awesome." Like really cool. Um, so it's short, takes an hour to read. Um, that's on Amazon. And then we've got integrated upstreamplanning.com as well as landing page for the book. Um, yeah. Cool. Anything else? You going to wrap it? No, it's all good. But no, I mean again, thanks you guys. Thank you guys for being here. Glad we could get you in while you're down here. So, while he migrated south for [laughter] for the for the week. Yeah. No, this was fun, guys. Appreciate it. Absolutely. Thanks so much.