0:00 All right guys we are we are back with another data drenched episode of energy vites here with my My trusty co-host Bobby Nealon. How are we doing today? What's going on? Do you get any any of your
0:16 stuff freeze over out in the country last night? Uh, no, it was uh, it was super easy in and out like coming back on Thanksgiving to hobby was insanely quiet and they were, I don't think customs
0:30 gave a shit. Like I could have brought back 10 gallons of tequila and they wouldn't have said anything, but yeah, I didn't
0:38 No, that's uh, it's as you can tell I'm bundled up because the heat in our building was not on, um Now that the weather's changed. So it's like 60 degrees. Oh, you're saying out in the country.
0:49 Not me coming back into the country. Yeah Nothing froze out here But that was a big deal of it in free so you would have been good. Yeah,
0:58 yeah, totally good Um, but yeah, no, no, no issues out here. Uh, I mean, this is what I imagine Calgary feels like in most of the July. So, yeah. So I'm looking at eight inches of snow.
1:14 Right now.
1:17 Yeah. Not quite the same here. We're complaining that it's in the thirties here. Um, or what is that zero? Yeah. That's right. That's right. Yeah. Oh, we're used to it. So that's all right
1:29 It's all good. Yeah. I've been waiting. We don't really ever get fall. We just get kind of a cool summer and then it goes right into winter normally. So that's kind of what we're in the middle of,
1:41 I think. But, um, anyway, we're here today with the, uh, the guys from energy payables, uh, Corey and Bruce want to give you guys a chance to, uh, to give us a little, tell us a little bit
1:56 more about yourselves Give you guys a second to intro yourself. Bruce Holman. I've been building software since I had a VIC-20 computer in the 1980s and moved from from that to mainframes to client
2:15 server to web apps and I just can't stop doing it. I've been in the old patch probably since about 1998 and it's a love affair for me I love the industry. I've
2:33 been out riding shotgun with pumpers.
2:38 I've been to some big facilities and sat with folks working there and I just, there's a ton of pride in the industry and I just, I love it. Yeah, so from my side,
2:52 let's see here, I graduated Oklahoma State I can oh seven and worked at Conoco Phillips. up in Bartlesville,
3:03 and I was in their accounting division. And so I started off in joint interest accounting. And it was around that time when people were still kind of accepting web-based applications, kind of the
3:16 dot com, invoice exchange. So obviously in the joint interest building side, we were, I was on the non-operated division And so we were looking at all the operators who were mailing their hard
3:30 copy jib and voices. And some companies were starting to use electronic means to distribute that and have their non-op partners code and enter all that data more electronically to reduce all that
3:42 administrative burden. And
3:46 so implemented a tool at the time called JibLink, which Bruce was one of the lead developer and co-founder for and.
3:58 Anyways, implemented that and actually joined the team with Red Dog Systems who owned Jiblink in late 2008. So Bruce and I worked together for a few years before he started a few other adventures.
4:11 And I was with Red Dog Systems up until 2019. We were acquired by Invaris and I worked with Invaris,
4:19 really calling the sales business development side for several years. And since then, you know, we've left and have joined forces with Bruce again. So we're kind of running the band back together,
4:32 if you will, which is pretty cool. Bruce really inspired me when I was in my early 20s. So getting to work with him is, you know, a real treat with that. And
4:44 so, yeah, it's just kind of been a passion, I guess, been the only guess, you know, for the last 16 years or so, you know,
4:55 really for the most part, it's all been around software So, thank you for your time. obviously within accounting, invoicing, trying to really help people solve these critical problems that have a
5:06 lot of overhead cost and administrative burden on how they go about it, which obviously has downstream effect if there's anything putting correctly into your accounting system. So being able to
5:17 really help the industry with that, leaving a legacy behind with the jibblink energy link application for many years and kind of moving on to the next adventure So it's just another chapter in the
5:30 book, but it's been a fun ride so far. That's cool. Yeah, I can't even imagine what the back office and everything looked like before, even like some of the software we have now existed. Like,
5:41 I mean, you see how big those teams are and just a number of transactions. Like, how was that even done? I don't, like, or how'd you close the books? I don't even know Our CEO cut his teeth in
5:55 the oil patch back office late '70s, early '80s, right? So it was all paper. And everyone's smoking in the office, and it was a different world. Just make it seem to be old data entry. Well, I
6:11 just remember where it clicked for me was, you know, we had a team of maybe 10 people, and this was Conoco Phillips, so I get that they'll have a little bit of a larger team, but you know, you
6:22 knew when operators would send their invoices. Like I could expect, you know, sell or BP to send us their invoice on the 16th of every month, right? And I remember one time, like it just totally
6:34 clicked for me that there has to be a better way to do things, and that's just kind of always stayed with me. So the person who would drop off the mail or the invoices around, we had this one
6:44 operator sending invoice, it was like 3, 500 wells on it, right? So there was no, hey, here's the envelope to unload. She would wheel, she would wheel the cart, this big part and it was just
6:57 stacked of it. And she would just leave it at my office and I'd walk out and it was like the most depressing morning on the 15th of every month because I'm like, oh, yeah, it's that time again.
7:08 And what am I going to do with all this? It was, oh, it just stuck with me, right? So I was like, yeah, anyway, we can think of technology, leverage technology, see where companies were like,
7:20 but that was like the moment for me. Like there has to be a better way like this. This is just difficult So were they still wearing green visors at the time or
7:29 just, I could just hear the accounting real low, the, the electronic calculators with the spool, yeah, yeah, yeah, well, it's why I didn't pack that. I can imagine people would be able to
7:42 work the books back then or where they were there.
7:46 It was out of control. I mean, there were what, six months to a year behind in processing, right, Corey? Yeah
7:55 Oh, it was insane. Now they're only three months behind. So we've made huge strides in 20 years. Yes.
8:03 No, let's unpack that a little bit though, 'cause I think there's a lot of people on the tech IT operations side that don't even realize the chaos that's happening behind the scenes once a field
8:17 invoice or a jib or a royalty statement gets distributed. Or they don't even know, like, oh yeah, we have all these fractional interests and all these wells and this is how they get distributed
8:29 and stuff. Walk people through kind of, you know, how y'all came about kind of building it and stuff and some of the
8:37 trials and tribulations along the way. I can, again, like Bobbi said, I can only imagine, you know, 20, 25 years ago what it must have been like to try and convince people that we're gonna do
8:47 this on a computer and stuff. Yeah, I mean, on the chip. front, these huge invoices that Corey's talking about, then it's real, right? It was incredible. And so for us, that was amazing
9:04 because it was, it was a blue ocean, right? I mean, GE had an old,
9:11 like an old product out. That was, you know, a couple of the big guys we're using, but really there was nothing there. Everyone was hand-bombing this into their systems I
9:23 mean, to actually get the source data and, you know, fan it out to the right partners, that wasn't terrible. But it really got interesting when
9:38 folks were saying, okay, I wanna dispute that, but I don't wanna stop the train, right? And then I wanna take that and pay it later. And that's where I think when,
9:53 When we came across that with our pilot group, it was just something we weren't expecting and we had to react really quickly and that was really the beginning of just getting into all the, I mean,
10:06 really into the weeds.
10:09 But I mean, any problem is solvable, right? And
10:14 I don't know, Corey, like so we started that, like we had a version 09 out in 2000 or so but I'd say by 2002 or
10:28 2003, things were moving along pretty good, right? I mean, we were still, we were still young in the industry. I don't know if we had even looked into the States yet. We were just working in
10:37 Canada at that point, but it was right around that time we started looking south. So what does that look like from a tech stack perspective? Like, are you, I mean, I assume the traditional way
10:49 was you just had a horde office.
10:53 accountants to some effects that are just manually entering the stuff and
11:00 then manually keying it in, or how are they doing it, and then how did y'all kind of build around that? Yeah, I mean, it was all, I mean, so working with IT departments too. So we're working
11:13 with the operators who are actually sending these things out And so
11:20 when you're starting a product like this, you also need,
11:25 like you need the right pricing structure and you need the right incentives for people to actually say, Hey, I want to try this, right? And so we took that operator data at no charge to them,
11:36 okay? So that was, of course they were going to be interested in that,
11:42 cut their mailing costs
11:44 and their shipping costs down drastically, and.
11:49 And so it was all data exports right out of the gate, okay. And like our first version was, you know, taking in these data exports and just, you know, parsing it out to the right, to the right
12:04 partners where those partners could generate a PDF. That was it, right? There was no coding. There was no, like you couldn't code it to your cost centers, to your AFEs, all these back office
12:19 accounts and so on. That all came a little later, right? But that was the start of it. So everyone was open to it. Yeah, there was very little resistance kind of. And it may have started,
12:33 Bruce and Calgary, maybe was a little more open to tech with that, you know, 'cause like this. Yeah, that's true. You know, well, jiblink at the time operator was an way only The wasn't.
12:47 distributing invoices to their partners. They were mailing them still, right? So, a worse case, if this internet thing, this product called jibbling doesn't work, we at least know that the
12:59 invoices are still getting to people the way they'll always have. So I think people always kind of felt comfortable that there was this little safety net in that regard. But I think once it started
13:07 in Calgary with it and a few people doing it, right? Like once a few could show that it was working,
13:14 you know, if it's working for Chevron up in Canada time, well, you know, someone like Chevron has the IT departments and kind of checks and balance. If it's going to work for them, it's going to
13:25 work for a small and medium guy when it comes to the technical concern, if you will, right? But that was, you know, and then I think what was really kind of helping you get down to the States is
13:36 you were saying, Bruce, is, you know, as they say, there's a lot of, there's a lot of dead skeletons Canadian border, whether people have taken stuff up from the US. to Canada to say it worked
13:48 here, it's going to have to work there, vice versa.
13:51 You know, and our approach was letting the Canadian market pull it down into the state. So actually, Chevron, Canada, called their US. office and said, Hey, you guys need to take a look at
14:02 this, get on board. Chevron was a huge flagship for us to get, you know, really kind of going into the lower 48 side, and that's how it kind of helped us get across the border a little safer.
14:13 And then, of course, everything kind of went from there. Can we take us real quick, step back, and for people who are listening, and can someone explain what a GIB is, you know, just so people
14:23 understand? Yeah, so for a GIB, a joint interest billing statement or a joint interest billing invoice, so operators of the well,
14:33 they have the obligation to take all of the, I kind of get the full timeline, right? There's all the suppliers and vendors in the field that that are performing work.
14:45 on whatever the project may be for that site. At the end of the month, right? So there's one site with accounts payable that has to process, receive all those vendor invoices, process and pay
14:59 those invoices for those suppliers. Then what's happening after that is the operator has to say, okay, here were all the charges tied to each site, whether those charges were capital charges to a
15:11 particular AFE, or if they were lease operating expense line items, they would break those out. And so you would then as the operator have to turn around and invoice all your joint venture, your
15:22 joint interest partners for what their respectable interest in the well was, right? So if I have 10, I'll receive the invoice, it'll tell me what the gross amount was, but if my proportionate
15:33 shares 10, then I owe 10 of what that total amount is, right? So it all starts with the activity in the field, Vendors, suppliers going to the operator. They got to pay those and then distribute
15:45 and allocate all those total expenses to their joint interest partners who have an interest or stake into that well. So joint interest and those invoices were distributed monthly very rarely would it
15:58 happen more than once a month, but it was just the once a month occurrence for that invoice for the partner to receive and then ultimately turn around and pay the operator for their share But again,
16:09 when I just think about, you know, say at GME, even as a running jib in an accounting system is a big deal and some of these like, if your accounting system isn't very good, it can totally choke
16:20 on itself and you might try to run it a handful of times it's, I mean, because, I mean, depending where you're at, you can have a ton of, you know, owners in the well and then multiply that
16:28 times all the transactions and it's I can get pretty unruly pretty quickly. It can, you know, and, you know, just as much as the other side, like, I know we've been talking about the joint
16:39 interest building side. I mean, at the same time. Revenue departments, Bobby, are having to administer payments to all their owners, right? Whether it's a working interest owner, you know, a
16:49 landowner, royalty owner, whatever it is. And the reason I bring that up is 'cause that's just even whole another ball game of complication, right? Just with, you know, are all the right people
16:59 getting paid? Are we paying them this month? Do they fall in minimum suspense, right? 'Cause their check amount's not enough. Did we miss any owners who should have been paid, right? Which is
17:09 obviously a conversation for a whole 'nother podcast for people, right? But just the breadth of individuals that have to be paid, there's so many more owners than there are joint venture partners,
17:20 right? So, but either way, it's still quite a feat for all the operators who have to get their jib clothes, their revenue clothes done in a timely fashion each month as you were saying, crossing
17:33 their fingers that their accounting system, their ERP system doesn't explode or delay things to get out on the timely basis all the phone calls start coming in. Yeah, or they can spend, you know,
17:45 50 million implementing SAP and that should work. That always solves the problem. That will always solve the problem. That's just the implementation. Right. That's not even the cost of the
17:57 application.
17:59 You get the coffee cup with that. You get three coffee cups with that, guys. Nobody ever got fired for, you know, procuring SAP, right? Or so I've been told. It's true Was Yael's vision for
18:14 this always to be kind of like a platform that everyone could kind of use? Or was it initially like, hey, this solves our problem for this client and then it kind of snowballed? Yeah. I mean,
18:27 the
18:32 vision was always capture the industry frankly where the where the real value is.
18:46 You know, every new operator that we would bring on would be a further cost reduction for every partner bringing down, you know, all those transactions that Corey was talking about. And it, I
18:54 would say the value of the product
18:58 to the industry, it kind of increased on a curve. The more producers that, you know, who are actually putting their joint interest building statements on there. Yeah, we joke, we used to say we
19:14 were shooting for a garbage monopoly, right? 'Cause it's a tough kind of grind part of the back office, right, doing that type of work. And yeah, I mean, it went great. When we were changing,
19:29 you know, Bruce, you've mentioned, you know, from GE back in the '80s, this really the only, there wasn't a jib at the time there was a revenue module I think. Actually, no, there was jib
19:40 there was jib yeah, but it was a it was a point-to-point solution Exactly, right? So I'd have to send it to Bobby and then I'd have to send something else separately to Bruce and since you know So
19:51 it was all this individual, but it was also built not realizing How many companies would actually exist in the marketplace? So what the dot-com kind of brought for Bruce and then was hey if we can
20:02 just take a source file from the operator and they have A thousand partners. Well now we can really stay to the thousand partners. And by the way instead of transmitting that From a point-to-point
20:12 thought where it's one at a time, you know, you just load one file Hit release and then the invoices go to the respective thousand partners out there, right? And they can all just log in and
20:22 access their information Obviously to just what is theirs. They don't see anybody else's so It kind of opened the internet kind of opened up that thought process versus the point-to-point. So we
20:32 kind of knew that going in, we were, you know. this was going to be built for the masses and to ultimately get there, you know, it was going to take time, but it's more beneficial for people to
20:44 say I'm logging into jib link or energy link, and I don't just see one of my operators invoices there. I see all 20 of my operators invoices there, so it makes it impactful from a, you know, full
20:57 breadth perspective. What was the
21:00 kind of unpack the evolution of your tech stack starting kind of from what you all started with, what it was when you had that bought out? I'd love to hear that.
21:13 We were doing, so it was Oracle. I mean, this is always our cloud, right? Well, it was a cloud, but it wasn't exactly sold as cloud. That's funny, because the first place we hosted at,
21:30 someone literally unplugged our database server,
21:35 Right, so there you go
21:38 And, yeah, it was, I mean, we, so when we started this thing Oracle was, you know, it was number one, it had the reputation for the capacity, that sort of thing. And so we leaned into that,
21:51 definitely. And we started with active server pages, and I mean, it was Oracle stored procedures, active server pages, and we had this kind of home-baked, kind of query management thing that
22:10 would sit between our app and our
22:13 database server. But
22:16 it was super simple,
22:19 and I mean, I do think there's beauty in that
22:23 But it was, you know, also, I
22:27 mean, it was,
22:30 it got the job done, right? It was rudimentary. I still remember our first login page It just looked like. Can I curse on here or is that a yeah, of course, it was dog shit completely and
22:48 But it I mean we evolved right so
22:52 We went pages back then we're pretty dog
23:02 shit Loading my my space my my space was pretty cool. Yeah, that's very Bobby I've always followed your my space. It's it's weird that you still keep it active, but that's great Me and Tom him and
23:14 Tom just inseparable. Yeah, no, I don't follow my
23:23 I mean we always Art sore front end kind of dev tools were always in the Microsoft so we Uh, we actually went to, I mean, we moved to dot net, but it was more VB dot net than C sharp, which is,
23:42 uh, I mean, hi, like everything we're doing now C sharp. Um, but, uh, that, that was, uh, um,
23:53 that, there were no huge refactors or, you know, big, you know, we're going to rip out this whole thing and putting something new Uh, I definitely got challenged, uh, on occasion, uh, some
24:03 of the young guys wanted, uh, sequel server, or, you know, another backend or something like that. And it's like, tell me why that would be worth it. Really? Like, just tell me. And then
24:13 we'll think about it. And there was never really an answer. So. Yeah, we joke about that. Yeah, it's like, what was going to be faster, but it had to be, you know, step change better, right,
24:23 for you. Exactly 100. Yeah, actually, so I'm
24:29 going back to you talk about active server pages. I'm assuming that's what ASPnet is.
24:34 And then to ASPNET, and yeah, that's right. So it was just right at the start of. NET, essentially. We don't need to go too deep in it, but like, what is, because like, what is ASP, all
24:47 right, active server pages versus like what people think of now? Like, did that contain like the HTML and some JavaScript? Or what was the deal there? Yeah, I mean, so
24:57 it was all essentially inline
25:03 Yeah, boy, that's going back, I'm trying to remember, but I mean,
25:09 it was all, you know, in-line processing, there were no,
25:17 I don't have a good answer for you, Bobby. It was
25:22 ancient, I mean, it was all, it was just HTML, And in this
25:28 ASP code, we weren't doing any JavaScript, nothing to start, and that all kind of came later. Oh yeah. That's crazy. Everything's like super JavaScripty now. I mean, I mean, it's something
25:39 you're doing a lot of like C-sharp stuff now, but like, you know, the leverage is in React or these different frameworks, but yeah. I'm still, I still have JavaScript allergies. I really do.
25:48 Yeah.
25:50 You know, I just, I can't stand that it's not strongly typed and that just kind of scares me and I will just. Oh, for sure. And then do my best to play. And
26:06 my best to play. And my best to play. And my best to play. And my best to play. And my best to play. And my best to play. And my best to play. And my best to play. And my best to play. And we
26:06 can jump back into some of that in a little bit, but like I think we need to jump to today. I mean, like so you guys are with energy payables. You know, what is energy payables? And you know,
26:16 what's, what are you trying to attack with that? And just, you know, let's just kind of dive into that problem set Energy Pables was formed about two and a half years ago. And the focus really is
26:29 around the full counts payable life cycle so is a single fully integrated platform with the vendor portal for vendors suppliers to submit their invoices and it's got kind of from the operator producer
26:43 side full invoice invoice workflow ticketing ticketing workflow and a pricing agreement
26:52 Module within the platform. So that way we can just tie everything together where I think what we're really trying to do is Bring better transparency. I think to the vendors and the producers With
27:05 where things are in the process are people Being invoiced correctly right the right amounts are things getting coded Directly right with that downstream effect. I was saying if it's all done
27:16 incorrectly it can impact, you know the company further downstream so I'm really trying to focus on that account's payable lifecycle.
27:26 There's obviously been systems around similar to like, you know, great. It's 25 years old, right? There's some systems out there that have been around just as long. Some people still do things
27:38 manual today. But we think kind of bringing, you know, a new modern tech step.
27:46 For folks, right, bringing kind of a line of customer service that I think goes back to Bruce and I with a Red Dog systems culture of really kind of setting the gold standard and the market with
27:57 focus, you know, heavily set on the client and the customer.
28:02 Really improving a lot of workflows within the app, the actual application.
28:09 So yeah, I think, you know, trying to do something a little different, a little disruptive to what's going out there. And again, just really focused on the accounts payable lifecycle. So would
28:19 you, would you say it is kind of comparable like an open invoice or wherever you see it. kind of in the market. Yeah, I'd say very comparable to Open Voice. I think with our customer base and,
28:32 It's probably. You've talked to who have used the Open Voice application.
28:37 You know, I've said a whole account like, not only are you guys doing everything we can do today, but you guys have obviously approached it a little differently. You guys had the framework and the
28:47 platform that is we maybe start asking questions around how can AI be implemented and some of the stuff or things you're able to probably evolve quicker within your product set.
28:59 So that's where I say it is. I think it's
29:02 also, for a lot of those users, no one's, especially they've been on open invoice for a long time,
29:08 there hasn't really been a major disrupter to the open invoice product.
29:14 And so it's not common for a lot of people to think that there would be something else or what is out there, but obviously a lot of people are asking questions. They just always wanna have their
29:23 finger on the pulse, what their options are in the market, who else is building something new that can evolve with our needs, regardless of the division or department at a company. So, I mean,
29:36 I'd say for the folks that have seen it, they've said, man, you guys are much further along than we probably would have ever set, and you guys are doing a lot of things that we would need in our
29:46 business today. So, I'd say it's very comparable to that. That's awesome Yeah, that one absolutely needs to be disrupted. So, I'm very excited to - I mean, I think in any market, there needs
30:00 to be healthy competition, right? Absolutely. I mean, there's been no reason for the platform to evolve or grow or be better.
30:09 So, at the very least, I mean, you guys should be pushing them forward, but hopefully also clawing out some market share from that. And then I'm trying to think with open invoice or something
30:19 like that, or like yours, like, it doesn't rely necessarily as heavily on like the network effects. Exactly, right? I mean, it doesn't. It doesn't like not like some other things. I mean, I
30:30 mean, I feel like the my gripe with open invoice from someone who's been in the services side my entire career is that it's all operator focused. And if I as the service provider want to get any
30:42 benefit out of it, I have to then pay open invoice way a bunch of money to integrate with their system, even if I am using a completely different system for everything else And that's it's not just
30:55 an open invoice problem. It's an industry problem where it's like, okay, well, if you use our platform and you pay half a million dollars a year for it, then you can. But the problem is as a
31:04 service provider, not all of my clients use open invoice. So you're going to like, how am I supposed to manage all of that? Like, I don't get the full benefit because the entire oil field isn't
31:15 using the same platform. And so I'm very curious on y'all's in how what kind of, you know, the way what y'all have done differently, what are some of the different value, value prompts you'll
31:27 have. John, I feel your pain and,
31:32 you know, I've had the experience as a service provider as well. And so we, we definitely, we came at this approach and design considering the
31:46 vendors as well. So for example, I mean, with our AI engine, any, like if you are loading, you know, invoices with any sort of manual effort,
32:00 you know, be it like a PDF or, you know, photo of a new voice or something like that, you can upload all that today. And we read it and we present it to you and say, this is what we read, how
32:13 we do it, right? And then if there's something missed, we'll, you know, we'll let the vendor fix it and off they go, they send it. Okay. And,
32:25 we don't charge the vendor anything, right? It's,
32:30 I don't know, philosophically, it almost feels a little disrespectful. 'Cause really, they're providing us inventory, you know, that we're passing on and making a buck off of. And so we'd much
32:42 rather just make their lives easy. And, you know, maybe spend a little time in our system, if nothing else, right?
32:51 We will, like, I guess another pain point from a vendor perspective is getting rejections back for stupid stuff. And, you know, we definitely, you know, we do a lot of validation before. We
33:08 show that validation before somebody hits send.
33:12 We're like, we're reading. So if there's a well reference or location reference or AFE number or whatever, account, anything on that, even if it's handwritten, you know later on, we're picking
33:23 all that up, we're saying, Hey. You know, we've seen that. We're pretty sure this is the AFE. You want to code this drilling invoice to you, right? And
33:32 so we're providing a lot of services like that as well. Yeah, that's my thing. It's like, you know, as someone who also builds software, you should pay for the software, but the structure of
33:46 the software that you pay for shouldn't be an all or nothing. And then if it is all or nothing, it only gets you 60 of the way there because the whole industry doesn't use it, right? Yeah, and
33:58 it's, I don't know, like there's some base functionality that we just want there. I mean, we talk about perhaps opening things up and providing some more analytics or something like that, but
34:10 frankly, we could also just give a nice API endpoint to the vendor and let them crush their data the way they want to, right? Rather than us try and think we're the smartest person
34:24 Um, so that, I mean, that is our, uh, I would say our philosophy, you know, it's, it's, it's your data kind of thing. You've loaded it as a vendor or you've received it as a customer. We're
34:36 not going to hold that hostage. And if you want to, if you want to bolt on a, uh, uh, I mean if,, if, if vendors want to send, uh, work tickets directly from whatever product they're using,
34:50 then great. Yeah That's, that's the whole thing, right? Like, that was always my issue is that field tickets, 999 of the time end up in a PDF. So who gives a shit? What system you're using to
35:05 generate said PDF? I'm not going to move off of that just to get a 60 solution for my field ticketing. Like, that doesn't make sense, but if I can give you a PDF and, you know, a list of clients
35:18 and all of their information and build like a mini, mini ERP, but with the extraction especially the way that you can pull stuff out of PDFs now like there's no reason like just like you said set up
35:32 an endpoint push push the data there let it extract and then map it's then it's a mapping exercise right like it doesn't matter and it's so much easier exactly right so yeah no that's awesome and
35:44 you're gonna i'm on a much smaller scale as an independent consultant i'd send out right handful of invoices a month but i was i opened anything when i was like oh wait i generated this awesome
35:54 beautiful invoice out of zero and i emailed it to them but now i have to go re-enter all of this by hand in open invoice in open invoice and then it's like oh yeah and just then you look up other
36:08 options and like well i'm way too small to pay that much for that this solution to we we had uh sorry uh job we we have one invoice uh one of the companies i was working at in the um But our customer
36:22 wanted us to split it against all of their active wells, okay? And so they had around 8, 000 active wells, but that would change every month,
36:33 right? And so, I mean, I actually started charging them time materials to load the invoice. So we just want to get rid of all that. So yeah. Yeah. Well, and again, removing the people from
36:47 like the people from doing the manual repetitive stuff, which is what computers were built for and are really good at.
36:57 And then just, yeah, you know, giving the users the features that they actually need to use it. That was always my like issue. It's like, oh, we've got an open invoice account. But now if you
37:07 want digital signature or digital approvals, you have to have this whole other thing. And like that doesn't work for a large majority of the industry. And so now we're back at square one, but all
37:17 the operators are using open invoice because because they're just giving it to them. And so it's, yeah, I'm glad y'all are taking a completely different approach to this because it really, like
37:27 once you have that field ticket, which again, is literally almost always in a PDF of some sort, you don't need anything else. Like, you know, it's again, it becomes a mapping exercise of just
37:40 data. Yeah, I think it's, you know, just a couple of things on it, right? You have, so again, our philosophy there is to be open source and play well in the sandbox, right? Whether it's the
37:50 vendor side, the operator side, right? We don't want to feel like we have to build everything, invent everything. If you want us to connect to systems you use that you're not going to get rid of
37:60 tomorrow, that not necessarily replace what's got to connect to it, right? We're not trying to create another module to sell to you and it has to be that way or the highway. So we'll connect the
38:09 way we need to. So I think just kind of be an agnostic is definitely approach that we're taking at energy payables.
38:19 you guys kind of been highlighting like on the pricing side. I mean, I think what's kind of fascinating about this is I think it's open invoice comparable. But I don't say that because I'd say I
38:29 think we even have some differentiators today that platform probably doesn't have.
38:36 But it's priced for everybody. So, you know, you can go up market and the application's gonna work, you know, for a large producer if they want it. But I could also go to a small, you know, a
38:47 mom pop operator that does 30 invoices a month and they can use it as well, right? But it's priced for everybody to be able to use it that works, you know, within them. There's no minimum entry
38:58 point or, hey, we love to use that platform. We just never been able to justify the cost. You know, that's not what we're trying to do with our pricing structure. So, you know, into that,
39:09 right? Like, it's one platform. What you get within the platform, the invoicing, the ticketing, the price agreement, It's all there, whether you choose to use it or not, it's there. It's not
39:19 like you're having to purchase and implement separate modules, right? That's the difference. So I think as the industry starts adopting energy payables more, I'd say we'll see probably better use
39:32 of the ticketing, better use of the pricing agreements. I think it's ultimately going to go to Bruce's vision around what we're trying to clean up with that transparency between the vendors and the
39:42 operators. And
39:45 you can just do so much more when everything's built from the ground up in a holistic platform than just individual products that. So
39:55 it's pretty exciting, I think, for folks. And there's going to be a lot of opportunity for people who need something like this, who haven't been able to afford it or justify the cost. And others
40:04 as well, who might be looking for
40:08 cost production and the time where oil is at the price it is, they're always evaluating what their vendor costs are, their GA costs, right? These are things that always just kind of come around.
40:19 I think the timing is right for us to be very open in how we've priced that structure for the entire market. Beautiful. So I mean, obviously you guys started this back in looks like 2023. So I
40:31 have to assume that the. Here real quick.
40:37 You have advances in AI and all this stuff. Like, what was that? What made this possible? Or have you always had your eye on this? And then it was like, all right, this is the right time to do
40:45 this
40:47 And then how has say AI helped, whether it's using it as a part of the platform, but also, I mean, as far as like the coding assistance and being able to build faster or do something like that,
40:56 just curious how that's impacted. Well, I mean, certainly reading the documents, we're not gonna write that from scratch, right? That's been done and it's great and it's evolving. And I mean,
41:10 we certainly have our own layer on top of that. So we're pulling out all the oil and gas components.
41:18 from an AI perspective, so I guess so two thoughts. One from the development side, and one from kind of customers or prospects talking to us about AI. What do you guys have and that sort of thing?
41:35 And just to follow up on that,
41:39 I am like, would you let AI manage your bank account? Right, maybe not yet, right? And so there's a level of confidence, right? And we're always talking about confidence when we build this
41:53 thing. And, you know, a lot of the work that we're doing right now, they are still, you know, they are non-AI algorithms, okay? And, you know, before we cut some of that stuff over, there's
42:12 definitely gonna be a bake-off and until there are conditions. that we can't handle today, that they handle everything we do today, any AI engine handles whatever we do today, plus more than,
42:27 yeah, we're open to that, 100. But we're not going to jump in blind at the cost of at the risk of data quality, right? Because it's all about confidence. And like Corey was saying, we want to
42:41 give these folks, the superintendent who's got 100 invoices to go through today, we want him to have super high confidence on what he's seeing and what we're presenting to him in our analysis of it.
42:55 And
42:59 that's got to be a controlled entry into the AI side of things, OK? Certainly for any AI function that we do, we present it to whoever requested that task and they get to check it over before they
43:14 move it on.
43:16 And it's great time saver, like, amazing. Don't get me wrong. And on the coding perspective,
43:24 like it's incredible. Really, I would say even, and the quality hasn't noticeably improved even in the last eight months. I will, I'll jump, you know, I'll jump between Grock and OpenAI. I'll
43:40 throw similar problems at both and see what kind of results I get back You know, we're,
43:47 like part of the way we can, we can operate this, you know, with really good margins is we keep our team lean and mean. And I think our dev team is, we can do a hell lot more with a smaller dev
44:01 team. No question. So, yeah. Yep. Let me, I got a couple of questions around that. Completely agree with you like people don't realize that giving a senior dev is gonna generally, speaking
44:18 nowadays, as long as they're using it, like they should, should generate senior dev quality code. Like it's incredible. And so, yeah, you have one dev sitting there with three different patents
44:32 going, working on three different projects. You get a very significant uptick in an output and review and speed and all of these things And so, I'm very bullish, like you are on this. We're gonna
44:48 be able to, we're already doing so much more with so much less. And so, like one of my big kind of statements or predictions at least, is that there's gonna be, you know, a sub-hundred person
44:60 billion dollar oil and gas company. And it's gonna be because they built it, you know, leveraging AI and automation and stuff. But I wanna go back to your point about, you know, a couple things,
45:11 one, when you're dealing with accounting. right? It has to be accurate. And so the way I always tell people, you know, that are interested in AI or all those things is it's like, these are
45:22 tools, right? Like they are not an LLM is not always going to be the best tool for the job. In fact, most of the time, it's not the best tool for the job in most situations. And so, but, you
45:35 know, where you can use it and like the, it sounds like you'll have kind of a human in the loop. When you're doing the data extraction piece, it sounds like you might be using vision models or OCR
45:46 combination of both. But it's like that part of it, even though it's not perfect, still gets you 80 of the way there, right? Like, it's almost instantly, right? Right. Exactly. To the user.
45:58 And so, it's like, just because it's not 9999 perfect all of the time doesn't mean it's not saving you 85, 90 of your time doing these things. And so, you know, I'm working on this project right
46:10 now where I'm having to normalize a bunch of frack data from a bunch of different frack data. providers across a bunch of different wells, and it's like, now I can give the language model the
46:20 headers of each of these tables and have it take its best guess at how I should normalize these things. And it gets me, you know, it saves me a shit ton of work and a shit ton of time. And so I
46:31 think that's another thing people, you know, have to understand is like, it doesn't have to be perfect to be useful. But at the same time, on the other side of that coin is where it needs to be
46:40 perfect, we have to take our time and make sure that it is actually perfect And there are a lot of times there's other tools that do better jobs that are already on the market for that explicit thing.
46:49 And so it's good to hear that from you guys. I think it's a balance, you know, like you need the people in there as well because someone still has to be accountable for that code, for that frack
47:02 analysis. And you know, to stand up for it, to explain it, to tweak it And I mean, so it's.
47:14 Uh, I mean, when I say that out loud, I suppose you could do the whole, these, these, these LLMs could probably possibly do the entire thing now, but I just, uh, um, I don't know. I, I
47:27 think that we are, you know, IT is a service industry and I still believe humans, you know, uh, can, can provide the right, uh, and kind of necessary touch for that Um, what kind of, uh,
47:45 what are some of your favorite kind of tools for, you know, extracting data from PDFs? If that sharing with, okay you're
47:52 you don't have to. If that's secret, so
47:57 yeah, well, we've, uh, I mean, we, so we, so I've worked on a couple of different. I'll rephrase that. So I'm not putting on the spot. Do you.
48:07 OCR based tools or are more of the vision plus OCR models starting to make their way. Yeah, it's more a vision plus, more a vision plus OCR. And like I said, we've got our own layer on top of
48:20 that. And so one, another project that I worked on
48:27 before energy payables,
48:30 we actually use named entity recognition and just through free form contracts at that. And you know, with the right training it was
48:44 pretty damn good. And we got really good success with that. And those were around like land ownership contracts, that sort of thing within the old patch. And like just to take a document and get
48:56 enough information to categorize it. Right, as you know, is this a taxing voice? Is this a contract? Is this a letter to a landowner? Right, that sort of thing that was pretty pretty cool.
49:07 Break NER down for the people who might not know. Yeah, so it's named entity recognition, which is a
49:18 not neural, but what's the word I'm looking for?
49:24 I'm terrible with all the acronyms right over the gate here, boys, but I mean, it's
49:32 almost like super pattern recognition, I would say, yeah, yeah, it basically looks at named entities across documents and builds relationships between them. There's all kinds of different stuff
49:45 you can do. It's really, really powerful, especially on like, that's the thing, a lot of the language model stuff, you know, we're, we do rag for enterprise. And so it's, you have all this,
49:56 it's a giant data problem, right? And people think, oh, you just give it to the, the LLM and it does its magic. And it's like, no, no, no, no, there's all this stuff happening before it
50:05 ever hit. The LLM is the last step and it's doing the least amount of work in all honesty, right? And so like we're, we're doing a bunch of NER stuff. There's even one that I've read this post
50:17 the other day. I thought it was kind of interesting but it was like inverted, NER. So instead of taking the most common, you take the most unique. And then that's another way of classifying
50:26 things, which I thought was kind of a interesting way of looking at things. Yeah, like I've never even thought about that, but that makes a ton of sense. But
50:34 no, that's really cool. I'm gonna flip over to Bobby, 'cause I know he's got some other questions, I'm sure. Yeah, no, I mean, just and,
50:43 I guess building on top of this, 'cause I mean, obviously you've had to evolve, you know, as software and everything evolved I mean,
50:52 you've done this probably full cloud, I'm assuming at this point, I mean,
50:57 I guess what is the preferred cloud? I mean, doing all the Microsoft I'm assuming Azure like later. Yeah, we're using Azure.
51:06 I mean, people really get religious about this sort of thing, and I'm not in any camp.
51:14 I mean, we're using pretty much the Microsoft stack So Azure just made sense. Makes sense, yeah. Yeah, well, that's the thing. The other thing there is like the industry just loves, the IT
51:27 staff and the industry loves Azure. And so that ultimately makes it a much easier decision out of the gate than having to convince people that, you know, GCP is the way that they should go. Yeah,
51:39 absolutely. If you tell them it's on Heroku, you know, they all get nervous, right? It's like the SAP comment earlier. I mean, Heroku's great, right? But it's, and you can do this a million
51:51 different ways. So we just went with, let's keep the debate low. And I mean,
51:59 like the team has worked on it in the past. So it's, yeah. Yeah, I mean, it all has everything you need. Absolutely. Yeah, they all generally do. And more.
52:08 But I guess with that, I guess on the AI side, are you, are you all utilizing the, like the data foundry or whatever the different stuff is within there for that side of it? Or is like, are you
52:17 trying to, are you big and stuff into the actual application or containerizing just. No, if so,
52:23 right now it's exclusively Foundry. But,
52:30 so, here's my pet peeve. You know, they've started charging by token, right? So it's like, whoa, what's a token? And so a token is essentially a byte of response data. So now we're,
52:45 it's like we're back to dial up internet charges and how much data did you use on your last session? And it just really feels like a visionless building model that just hampers creativity. And back
52:59 when I was doing mainframe stuff in the early '90s, we had to be careful how many bytes we used when we were storing information. And I thought those days were done and now I'm seeing tokens. So
53:15 when I see stuff like that, Bobby, now I'm thinking, yeah, Maybe we should be just taking a model, rolling our own. and staying away from something like that, right? Yeah, and I wish I didn't
53:27 have to make that decision. No, yeah. I think the token model, you know, it makes sense for the big providers, right? Like that's what they want. Yeah, absolutely. Is they want one and then
53:37 they also, you know, GPT now has the routing layer. So it's like, oh yeah, I'm gonna route you to the best answer, which probably also has the most tokens just for your pleasure, even though
53:48 you didn't necessarily ask for that
53:51 But the flip side of that is like, I completely agree with you, like as an enterprise, they want predictable, forecastable pricing. They don't want, oh, well, next month, it's all gonna
54:03 depend. Like that doesn't work for the enterprise users. And so I'm also very curious to see how this shakes out in the longterm, you know, like on a consumer perspective, like me using this for
54:13 one little thing, okay, fine, I'll pay my 50 cents for my query. I don't care. at an enterprise level. And when you're building tools on top of those tools for enterprise,
54:24 it's different. So I'm curious to see if everyone just like hard forks and ends up with their own models that they're hosted. Like are we going back to self-hosting everything? Or are we gonna find
54:35 a different methodology here? Or are we gonna roll out ads in the middle of our responses so that the costs are covered? That's absolutely coming. Yeah, that's funny to say that. Yeah, I agree
54:48 But I mean, yeah, like I'm thinking about the hard fork. And guess what? It's more controllable. You don't have to apply for a quota. That's another beautiful benefit of that. I don't think the
55:02 average person realizes like how hamstrung every single cloud provider currently is with GPUs and hardware controls. Exactly. Whether it's a huge risk to you. I mean, you're providing a service
55:15 that people expect to work in a certain way if something out of your control happens, that they're like, sorry, you're out of tokens, or you're on it, and your clients are screwed for the next 10
55:26 days until you reset. Or we give them a user experience that lets them bang away at this LLM, and their usage goes through the roof, and we're charging them a flat fee, and now we're hurting
55:43 ourselves, right? Underwater, yeah You got to manage the op costs and this per use thing is, it's just a, it's an irritant. I mean, the tools are still beautiful, no question. And now they're
55:56 talking about tune instead of JSON responses, right? So it's basically CSV files all over the game, which is another, like I hate CSV's.
56:06 Corey, you remember CSV's? Oh yeah. I mean, I still remember them. They're still there. Yeah, they're still there. Exactly. They're rocking They're not strongly typed, are they, Bruce? Um,
56:18 I saw what you did there. Yeah, no, they're not, but I can, I can read it, but, you know, at least Parquet has types. Um, but I digress. Yeah.
56:28 There was a joke article that came out a couple of weeks ago and it was like SCV beats tune and it's literally just separated comma values.
56:39 Yeah. Yep. That was Well, and I'm, I'm more a fan of like till the delimited five, but, um, but it's, it, it brings up a good point though, because there are people doing really interesting
56:54 things around token efficiency, I guess is the way to say that, right? Like, uh, I saw something I want to say it, they, they took a picture, like a P and G and a comp of the tokens of the
57:09 prompt of the query, whatever compressed it and and fed it to the model that way, and saved. some ungodly amount of tokens just by doing that and so I'm very curious, it's interesting from that
57:21 perspective because it's like, okay, now people are trying to come up with all these workarounds around
57:27 the constraints of the business models that they're forcing upon us. Necessity is eventually different. It's not terribly different than what, five, eight years ago, like with cloud spend and
57:40 people, there were services literally that would manage spot instances for you. And like, you know, you didn't have to know,
57:50 but it would bounce around and it would find spot instances that were available to run your workloads. And it was saving a lot of money, but you know, you had someone either that to be a layer for
58:02 someone to manage that. Because otherwise you're paying on demand or you had to reserve instances if you actually knew how much compute you were going to use. But I guess all that's going to come
58:05 back around. Anything that's consumption based, someone's going to find a way to help hopefully Hopefully you try to mitigate it as much as possible. Yeah, game it. Yeah, I mean, there's even
58:14 layers on top of like, there's this company, I won't mention them now until I get them on the podcast, but they're like a middle layer between, say, Snowflake and like the BI layer. And
58:25 basically it identifies if it can run that query locally with DuckDB, you know,
58:31 and most of them are small enough queries to be executed by DuckDB, you
58:35 know, against the data. And I don't know if they have something to cache or how they're doing it. And then it'll only send it to Snowflake to run the query, you know, when it's, you know, above
58:44 a certain size or where, you know, you actually need a Snowflake source power, but, you know, that stuff's all coming. And all that stuff makes me nervous as an architect, 'cause like the more
58:56 links in the chain, the more things that can break, right? So - Points of fire, yeah. Yeah, 100. 100. I
59:05 mean, I guess the last thing, you know, before we maybe get into the speed round though, is like, because we were talking about it earlier, I didn't really get to it. in API endpoint in, say,
59:14 the energy payables, but like, what do you guys have now or what are you envisioning as far as like sharing data out of the platform because I can speak for the incumbent that their API is really
59:25 difficult to navigate and there's entire companies built off of being able to navigate that API and integrate it with other stuff. So I mean, have you even much thought to that side of it too? Yeah,
59:35 yeah, absolutely So a good buddy of mine that
59:43 I worked with at Muddy Boots, he implemented GraphQL,
59:49 which is beautiful, right? And I definitely think that is the nicest way. I mean, it's early days for us, our hair is on fire, you know, just within the current app. So we're not doing that
1:00:06 yet So right now it's kind of standard. uh, you know, just kind of standard REST APIs, uh, um, but we're, I mean, we're doing things like, uh, you know, go pull your work tickets every day
1:00:20 so you can do your daily accruals, right? I mean, that, that, that API is just there and you can do that. Uh, with, again, as just part of the, uh, as part of the offering, okay? Um, we,
1:00:33 we have a lot of report, well, we have reporting, uh, within the product, and, uh, you know, we just have, uh, you know, we just have exports from that as well within the app. So, you
1:00:43 know, someone doesn't want to work with, you know, their IT department, they just need the data. That's all there as well. But I mean, right now we're, we just have a suite of rest APIs that
1:00:58 we're building on. Nice. Yeah, because I know, like, obviously some people, and it'll probably depend on, you know, if you land some of these customers, really are big enough and are going to
1:01:07 make it worth your while to do it. But like, you know, whether it's snowflake data shares or stuff like that, a pretty big deal too. Because yeah, getting a SQL database of it. Basically,
1:01:16 there's a company that exists for that reason to be able to create, you know, integrations and create a SQL database of your open invoice, you know, you know, data, because just it is unruly,
1:01:26 unless you've got a lot of experience with it to get the data out reliably. Mm hmm. And yeah, I mean, we will, we will and it's an interesting discussion, right? Because a lot, they're
1:01:36 essentially getting all of this transact transaction data into their infrastructure already. So they have a source within within their house today, as well, right? So, yeah, it'll be, we're
1:01:52 open to the conversation. I'm looking forward to it. Right. But I mean, I mean, you have to focus on, right? Yeah, exactly.
1:02:01 Yeah, no, I think GraphQL is pretty great as well. And I guess I want to go to depart on the rabbit hole because like rest, basically, you have this endpoint, you have to pull it, you'd have to
1:02:09 pull like five different points and then you'd have to join this up together. But with the GraphQL, you can kind of write your your API request that pulls or multiple endpoints and consolidates it
1:02:17 into one response, correct? I mean, oh, it's a beauty. Yeah. And it's a friendly query language. Right. It's yeah. Yep. That's definitely the preferred destination for us. Beautiful John,
1:02:32 you guys want to jump into a real quick speed round, like one or two questions for guy? Yeah, this blue by guys.
1:02:38 We love love. It always does and we're nerding out over this stuff.
1:02:45 This one is for Bruce. What what is either your favorite or most kind of interesting, exciting It's a tool that you've come across in the last six months that you're using, whether it's a Python
1:03:02 library or web service. anything like that. Oh, I'm terrible at thinking of my feet, guys. It's
1:03:09 okay. Yeah.
1:03:12 I'll pivot to what, where are you located? Are you in, you're in Canada, yeah? Calgary. Okay, where do I need to hit up when I come to Calgary next with, I'm gonna bring my wife next time I go
1:03:27 up there 'cause it's so awesome. Have you been to Caesars? Have you been to Caesars Steakhouse? Corrino Caesars,
1:03:36 Caesars is old school, wood paneled. Oh yeah, that's what I thought about it. You know, white shirts, dark pants, all the serving staff is top shelf and it's literally the best steak you'll
1:03:50 find ever. I think the last time we had someone from Calgary, I think we may have got similar answers, so. I think you're right. Caesars, yeah, it's describing it.
1:04:03 Beautiful. Free ad campaign for seniors. Yeah. Second what the, they don't need it. Yeah. Sticking with the food theme, Corey, where's the favorite place to take clients to lunch in the
1:04:16 Houston area? Oh man. There's not many options to choose from, I know, but now it's bad. Well, I guess, you know, for people outside of Houston, right, it always depends on where the client
1:04:28 is located, right, 'cause that'll - Houston isn't our way for - That'll be for starters Yeah, man, like, just even kind of an OG spot that I'd love to take people to is the LTM Po on Washington.
1:04:33 Always, even for lunch, had kind of a good vibe. Man, I'm trying to think a little centrally located, but that usually was my go-to, 'cause I knew I could do it great for lunch or for
1:04:48 dinner. All right, what are you ordering at that LTM Po? So,
1:05:01 I like spicy food, so I gotta test it, but I always love to get the salsa off the menu. Oh. Like all the mocha hette, but really mocha hette is just a, actually, I think a bowl that you put the
1:05:11 salsa in the corner. Yeah, I grind it up in there. Yeah, so I start with that. Then always, you gotta get a pound of
1:05:21 the fillet and then I always tell them to put some grilled shrimp on the side, 'cause we said they have to have their little hot tub and the butter sauce there.
1:05:30 Listen, the butter hot tub is so underrated. It's like all the places here do it, but then you go outside and no one has it. It's like, guys, this is so simple. It immediately makes the food
1:05:41 taste better. Just give me a tub of butter. Yeah, you gotta walk by my table and understand it's a judge-free zone on how everything is set up there. So, yeah, that's usually gonna be the go-to
1:05:54 set up. Yeah, fajitas from El Tienpo are always a good option. Yeah, Bruce, ketchup flavored chips, year, nay.
1:06:04 He's in. I'm seeing that make its way down to the States a little bit more. I mean, I remember going up there 15 years ago, Bruce, and you guys were talking about ketchup chips, I'm like, what
1:06:12 the hell are these things, right? Oh buddy. And now like Trader Joe's has ketchup chips, Lays is doing their thing. Takes us a while to catch up. Yeah, I got a care package from Buddy up in
1:06:24 Calgary a couple years ago and had those along with few other things But that was like, all right, this is different, but I'm gonna go for it. That's pretty good. My wife's favorite.
1:06:36 They're so tasty. They're unexpectedly good. Oh yeah. Yeah, well, love it. Well, cool guys, this has been awesome. We blew through our time today. Where can people reach out and get in touch
1:06:49 with you guys if they wanna follow up? What's the best way for them to do that? Cori Yeah, so CoriEnergyPablescom.
1:06:60 I've got a LinkedIn page as well there for Corey Scott and then, of course, the Energy Pables website just has some information there and just if people are interested to reach out through that,
1:07:11 everything gets directed towards me. So
1:07:15 best point of contact, those are probably the best avenues. No one's probably trying to hit me up on TikTok or anything else. So I'll leave it to those areas. Hey, John, got some good run on
1:07:25 TikTok about a month or two ago Yeah, one of our clips hit like
1:07:31 600k on TikTok. Neither of us, I don't have a TikTok, but the company paid for this. So yeah, it's like this is awesome. I get all of the benefit and none of the repercussions of having to read
1:07:42 the ridiculous trolls in the comments. So troll me all you want because I'm not reading all of you. Thank you to all the idiots who left their credentials in their GitHub repos. Yes. Keep posting
1:07:53 your API keys to your public GitHub accounts Oh my gosh. That's awesome. I found some ridiculous ones. You're kidding. No, like, heavy SaaS tokens and, you know, enterprise platform tokens
1:08:09 that are just there. But yeah, no, the tech community loves to talk about how you just like the oil field, about how they're the smartest ones in the room. And so any kind of video that we put
1:08:21 out there that has anything mildly controversial, they like to chime in and talk about it or give their their their take, right? They were like, this isn't a new trick. Everyone's been doing this
1:08:32 for 20 years. That doesn't mean it's still not happening and people still don't know about it. Anyway, you want to wrap us up? We appreciate you guys having both of us on, man. Yeah. Yeah. I'm
1:08:43 glad we get the time out. And yeah, I mean, I wish you guys the best and yeah, hopefully people are reaching out after this and we want you guys to go disrupt the industry. So it's good. Yes
1:08:56 Keep fighting the good fight.
1:08:59 Oh, and your data. Appreciate it, guys.