Energy Bytes

0:00 We are back guys with another episode. I know we've had a break. We appreciate you guys bearing with us through the holidays and everything we are getting more consistent. We have a lot more lined

0:10 up on the calendar for you guys in the coming weeks, so we appreciate it. Today, we'll just jump right in. Bobby, how you doing, man? Doing great. Happy to be here. And then today's guest,

0:20 we have Vicki Nott,

0:24 the co-founder and CEO of CruxOCM. Yes, thanks for having me Thanks for joining us, and we appreciate it. Are you calling in from the great beyond of North? Yeah, today I'm in California, so

0:41 I'm in the land of the free.

0:46 Very nice. Enjoying the weather? Oh yeah, much nicer than North. Is everybody? Oh, sorry, so your in San Francisco, is that somewhere you are often, or that you live out of there or. Yeah,

0:60 here about 80 of the time and the reason why is like yes, our customers are oil and gas here at Crux, but we're a software company and we're the best place in the world to build a software company,

1:12 it's California. So we've got lots of presents in Houston and in Calgary of course, but yeah, we spend, I spend a significant time in California and like our talent, you know, our product and

1:22 software engineering talent, especially executives. Yeah, we came here to get those folks. That's awesome I wish more people would recognize that the talent isn't in Houston for software

1:39 development. It's not, unfortunately. Hopefully, but on other sides of it, I mean, then like Houston has some, but then like I'm seeing people trying to hire like high-powered data people in

1:45 Midland, or even try to get them to, or you know, Oklahoma City is not as even dire as that, but it's like, you realize what you're gonna have to pay someone who's really good, if you can pay

1:54 them enough to get them there, you know, like Yeah, and just to find them. Yeah, it's, they don't, yeah, folks, folks that love that caliber of skill set in the tech industry, they

2:07 definitely prioritize also, you know, lifestyle and, and yeah, getting them to move to Midland would be, I would think almost impossible for that type of talent. Yeah, like, I don't know, 500,

2:17 000 years, not enough, and like, no one wants to pay that person that, like, for what they need. Right. Yeah. No, and that's the thing to you that I definitely find with the talent side of it

2:26 is, yeah, lots of folks, as we were chatting earlier can write some scripts Think they can be a software engineer, not understanding that, you know, how mechanical engineering is a discipline,

2:35 software engineering is a full discipline. Like, it's unfortunately not that simple. That's what I definitely learned switching over into from back in the day I came from a pipeline. So I trained

2:45 as a control room operator for the Keystone pipeline for eight months and did about 12 hour shifts. Yeah. So coming from that, you know, being a chemical engineer, I was like, oh, I can

2:53 definitely write software. I can write a script. I can do mathematics mathematics of what I mean.

3:05 Now that we're great and then what I've learned since starting a software company is that that is a whole other industry and you know if you're trying to solve a problem you got to write a script

3:13 that's great but if you need a product that needs to be maintained it's it's another it's another level of intensity I guess we'll say How has it been in Silicon Valley would say like have you found

3:24 that the talent is good enough of I mean I'm assuming the answer is yes but good enough to justify paying Silicon Valley you know salaries and benefits like executive wise. Best talent I found from

3:36 like but like for talking like you know software like product management talent like software engineering talent like. Yeah definitely hands down like our executives are here in the Bay Area

3:45 individual contributor talent though for folks like this is probably would be interesting Canada has fantastic individual contributor talent at a fraction of the price of California. So if they're in

3:56 their Canadians they're awesome. And they're lovely. They're super nice. They're great. So we hire a lot of our software engineers actually from Vancouver. Okay, nice. What is, I mean, this

4:07 kind of brings them up. What is it about Canada? Is there something, whether it's regulatory or wise or whatever, like, Canada seems to pump out really good software, but also really, like,

4:14 people on your side of the world, like, industrial process controls in Canada, like, you know, IOT type stuff, like, I mean, I've seen really great companies. I mean, is Cold War out of

4:24 Canada as well, right? Yeah, I mean, there was another company, I can't remember their name now, but John and I worked at a company that would be kind of in that realm, that reservoir data

4:33 systems. I remember we saw, there was a company, and I was like, looking at what they could do, I was like, I was praying to God they didn't come down south of the border, because, like, it

4:41 was like, man, these guys look like they know what they're doing, but, and I think maybe, like, there's certain requirements of, like, monitoring that go on, say, on the well side that, you

4:45 know, require that, but it seems like Canada pumps out a ton of talent in that space Yeah, it's also true because like um

4:55 Canadian like physical engineering, like chemical, mechanical, electrical, it's all really, really regulated. So like when we, like, and the programs are the same across the country. So like,

5:06 it didn't really matter that I went to University of New Brunswick, I got the same education pretty much as McGill, like it didn't, that's a country difference that I think. So I think like the

5:14 standard of like that type of engineering just happens to be really high in the country because it's been established for so long and regulated for so long. Like we're in a position that when we

5:23 graduate from chemical engineering, like I could write your guys' PE exam without studying. Like that's where they try to get us to.

5:31 Do you have your program on? I do, yeah. That's the other fun fact about Canadian engineers they all have is a silver ring. It is a stainless steel,

5:44 but it's, you know, not a cheap ring overall if you, if you think about it. But yeah, it's the whole point of it is, and I think it's quite lovely actually.

5:55 when you're signing off on your critical settings or your critical designs that can impact the public, your ring touches the paper and you're reminded how important it is. I'm like, that's just a

6:04 lovely thing. And I think it's very true in the IoT skata space, right? Like these things move multi-billion dollar critical assets. You have to treat them with utmost respect. Absolutely. All

6:15 right, so do you get your stainless steel ring and you come out like, how does Vicki get from that to sitting in a control room so where we're at, it doesn't have to be super detailed, but what

6:27 was your journey like? So it started 'cause I started in labor and pulp and paper. And I didn't even know what an engineer was at the time 'cause I'm from like, Outport, Newfoundland. And yeah,

6:37 so I started in pulp and paper, labor on machines. Paper mill. Paper mill, yeah, 12 hour night shifts. That's awesome. There's so many engineers that started out at paper mill. It's amazing.

6:48 And it's a fabulous place to learn. Like it is intense, but intense crazy you like. things catch on fire and like, it's a, I loved it. And so, but having like worked labor on the machines when

6:60 I went into oil and gas, you know, I was like, okay, I need to, really naive of me. I was like, I need to be able to operate this thing if I'm gonna be an engineer. Like you can't expect me to

7:08 be any good at this job if I don't know how it operates. Like that was just my, I just didn't understand that engineers usually work on things without actually operating them. Like I didn't know

7:17 that was a thing. Which is a terrible thing in all, like in practice, right? It's like, show me something that's hard to take apart and I'll show you an engineer who never touched the thing that

7:28 they designed. Like that's why I became an engineer is because I was sick of how poorly designed my car was in high school. Just the simplest stuff. And you're like, if the person would have done

7:38 this one time they would not have designed it this way, right? If they even sat in the car, they would have known, right? Yeah, no, absolutely. Yep, exactly that. Yeah, so you have to came

7:48 into oil and gas thinking like, Okay, I need to be able to operate. It took me actually a year to convince the control room manager to let me operate Keystone as a trainee, right? I was never

7:57 ocued for making sure that that's super and I was under supervision and all of that. You know, safety is the most important. But yeah, so did it for eight months. Realize that the most modern

8:09 pipeline control system in the world is antiquated compared to a hundred-year-old pulp and paper now.

8:14 That's that's like the craziest part of our industry is the minute you get outside of it. You're like, what the hell are we doing? It's just, you're like this hurts. Yeah, there's so much tech,

8:25 like we think of tech debt from a software side, but there is a ton of tech debt from a hardware side as well. Like, and it's, yeah, it's terrifying. Well, the OT tech debt, it's spectacular,

8:35 right? Yeah. Yeah. It's like, there's no connectivity, there's no way to even, like it's all created as like, you know, you install it on your server with your floppy disks. Right, well,

8:45 and it's like architected that way too. It's architected that way, yeah it's not like so it's not actually. That's what I think a lot of people miss is like with the current OT systems, yeah, we

8:53 can we can do these cool connections and it sounds all fantastic, but they're one-offs and you can't connect the whole system because nothing's designed to be to be done that way. It's intentionally

9:02 designed

9:05 by the Purdue model, which we've now realized is not as safe as we like because there are many on-prem cyber security breaches. So really the zero trust environment is actually the safest, which is

9:15 probably where we need to go Not probably. I think it's absolutely where we need to go, which is then, you know, trusting the Microsofts and the fact that they've got the keys to the castle. Like,

9:24 that's OK, right? You still own the keys. Well, certain things that they're just going to do better. And I mean, I've always argued that like in the cloud, they give you every lever that you

9:32 need to secure. I mean, but again, there's the shared responsibility model and you need to, you know, secure it and apply best practices. But because it's not a good standard But just because

9:43 it's like sitting in your closet doesn't mean that that server is more secure just because you can go touch it like it actually doesn't.

9:50 Yeah, right, just 'cause you can touch, it doesn't mean it's more secure. And that's, I've learned a lot about that 'cause like when I first started, you know, like I was coming from oil and

9:58 gas company industry, I was like everything must stay on prem. And then as soon as I started understanding like how limiting that is in terms of like general functionality and then how unbelievably

10:09 unsafe it's becoming like over time, I was just like, whoa, like why are we, why are we taking this as a gold standard? Like we, sometimes you need to throw out what you did in the past and

10:19 that's okay. Well, I mean, 'cause then at that point, then you get into like physical security. I mean, you've got to have, you know, controls on who can get into the server room or, you know,

10:28 or anything that goes on. And John and I have talked about this, like especially on the upstream side. You can walk out to most people's well-sized facilities. There's no big, you know, barbed

10:36 wire fence around it. That's keeping people out. Or you can just cut the fence with some work, Edders. Yeah, if you can just put it up and open the cabinet and it's, there's the password one,

10:46 two, three, you know, right inside the, you know, the box and like, And somebody probably left their thumb drive there with the last PLC updates and. Yup, yeah, and are you updating the

10:58 firmware or whatever else, I mean, it's just and then I think I'm gonna expect your nightmare. In fact, your point about integrating. I mean, I'm, well, by the time this comes out, I will be

11:08 done with my time at my current employer

11:13 where we're integrating, but like those, they're not getting that skated, it's our skated system from GME integrated fully till a year from now I mean, like they still have to run them in parallel.

11:20 I mean, because, I mean, I think we had a really good one and I think there were certain things that we were doing that, you know, my current employer wants to get to, you know, but at the same

11:30 time, like, you need to have all that integrate with the other assets that are adjacent or even just around, you know, and how do you get that data into the company, you know, data repositories

11:39 to expose it. It's just, it's kind of a mess. That's where I like where Microsoft's going to be honest 'Cause like, they are. the OG software company, right? Like they understand networking,

11:53 they understand platforms, they understand all of this stuff. And so, so I think like in terms of making that migration to data connectivity, like their fabric product sounds perfect to be honest,

12:03 'cause it's like it can go across all of them and take everything out. Like we don't have that because like, you know, Pi process book doesn't talk to anything, right? So you need another layer

12:12 in there, unfortunately, whether we're like it or not. Well, it's always like been baffling to me and maybe it's just 'cause I'm coming into it later, you know, after maybe this solution was

12:20 necessary, but like all these people have Pi to talk to Signet. But then we're gonna get the data out of Pi to put it into snowflake now. And it's like, why can't we just go like straight? All

12:30 right, why do you, you know, I'm assuming Pi served things that SQL Server oracle weren't serving at the time, but it's like - But now it's not making a lot of sense because the database and

12:41 capabilities of Pi are so rudimentary that like, Cracks our software already just does the exact same thing. So it's like, okay, well, we already have all that data. If you want it, right,

12:51 like you can go around now. But that's, but we're one of the only solutions that are enabling that, right? Like all, all of our, all of our employers, all of our customer, like, you know,

13:02 all of the oil and gas operators, they're trapped. Yeah, that's the, that's the like recurring theme of the last decade in oil and gas software across the board, regardless of what side of it

13:14 you're on It's like, oh, hey, this product sucks now. Someone's built a new one, but all the data is still trapped in the old product and they intentionally built it, so you could never export

13:27 it, never get it out, you know, make it as difficult as possible to get anything out of it, and now they're just trapped in it. And it's like, what a great customer experience to feel like

13:36 you're trapped in a tool that you're, you don't like, like, that sounds awful. Yeah, oh, and it's like, you know, it's just the evolution of things, right? like, pie from the 70s, 80s,

13:47 90s, like - was like the best tool there is, right? And that's fantastic. And like when I first started on pulp and paper, I loved using it because I was like, oh, this is amazing. I can pull

13:54 whatever I want. I can query everything. But now it's like, okay, when you look at LLMs and you look at like what your phone bank can do, why, why are we still doing things the same way in oil

14:07 and gas, like we can do better? No doubt. So one thing I can maybe articulate, and I mean, maybe this gets into what you all do as well, but like, what's the difference between industrial IOT

14:21 and SCADA? I mean, obviously the lines get blurred, I think, but I mean, for me IOT industrial, gets from like, I've got a sensor or some kind of data acquisition device and it goes right from

14:29 there to the cloud or to some repository, whereas like the SCADA, like it's part of this entrenched kind of network. And then they may have similar components that, but just I'd love to get your

14:41 thoughts. So I wouldn't go straight to the cloud with any end device from a control perspective. And the reason why - What you're saying is that the controls piece is the great delta. Yeah, you've

14:49 never came out from the controls, like I don't like that at all. And the reason why is like if you think about it, like you've got like an analogy, you've got this really, really dangerous thing

14:58 that you've got a fence around, right? And that first fence is your PLC, right? And then that second fence is SCADA. Why would you purposely tear down those fences around this unsafe thing and go

15:07 straight in when you can nicely go through the fences and make sure that and put up another fence and stay super safe? So going straight to the end of ice, I think is not a good idea. I don't think

15:18 SCADA will become obsolete at all. I actually think it's an extremely necessary layer. I just think that it can be, yeah, made significantly better. Okay, so, and I'm assuming is that what crux

15:29 is solving for the midstream space? Yeah, so we are a layer, we're an advanced process control layer on top of SCADA. that enables an autopilot or an adaptive cruise control functionality for

15:39 control room operators. And we've designed for the control room operator instead of designing for the advanced process control engineer, like we still designed for them because they're very, very

15:47 important administrators. However, they're not actually the end user, right? And so all advanced control systems that have been designed for advanced process control engineers, I believe they're

15:56 leaving extra throughput on the table because your utilization's only 50 to 60 on those APC packages. Whereas what Crex is seeing on pipelines is 70 to 80 utilization, so we're seeing even more

16:06 throughput than regular APC packages.

16:10 So is that how you got the - I mean, is that what made you jump in and start the company? Was you saw this giant glaring gap that everyone had and no one had filled yet? Yeah, exactly. I was just

16:21 like, this technology exists. It's proven in every other heavy industry. There is zero reason why this wouldn't work here. Like, let's give it a try. And me and my co-founder, we mocked it up

16:30 in Malav with some amazing historical data that we'd had And we put it - It works, we're like, holy crap. We're like this thing, like advanced control on a pipeline, like really, really works.

16:42 Well, it's amazing to like, to your point, when you have good data, when you have a bunch of data and good data sets, it's amazing what you can do with automation and it's like, we as an

16:53 industry, either one, don't have good data sets, we think we do, but then we get into them and we're like, oh, these are awful. Or we've got a ton of the data and we're just not doing anything

17:03 with it, which is also crazy It's like you're paying money to acquire the data, to store the data, to stream the data, and then it's just sitting on the database somewhere and you're not actually

17:13 using it. 'Cause a lot of people don't know how. Like, they actually don't know how to query it, they don't know how to understand it and look at it. It can be overwhelming, right? It's a ton

17:22 of data, right? Like, Bobby and I - You don't know where to start, right? Exactly, when you work in the sensor level, when you even, you know, have one second intervals or even higher once

17:31 you get into the Hertz range, It becomes big, it feels like it's big data to a human very, very quickly, right? You only need a few weeks of one second data for it to not fit on a plot anymore,

17:43 right? Like, and so. Yeah, and then you need to be able to abstract it yourself and not everyone has those skills, right? Yeah.

17:50 OK, so you guys have been around what eight years now looks like? Yeah, too long. But the first two years were like, I didn't ever heard of Silicon Valley. I'd never heard of venture capital.

18:01 I'd never heard of startups, really So first two years, we got nowhere. Let's not count those. No, fair enough. But like, so I'm still trying to necessarily understand, you know, exactly

18:12 where you guys are. So you know, people have like their skater implementation, you know, and that's can, you know, where they can control the pipelines, they can do those things. But so where

18:20 do you fit in, like, and what, what were people doing before? And now where does the software sit? Like, yeah, so

18:28 I'll go with like the keystone example, because it's the easiest one. Starting at Keystone takes 1, 100 commands, like 1, 100, takes about three hours to do, and that pipe one shuts down on

18:37 average two to three times a week. And this is pretty normal for pipelines all across the industry, like they kind of operate that way. You put on the pipe bot software, so what you're doing too,

18:46 is you're manipulating all of your pressure set points, so your suction pressure set points to maximize your flow, your throughput, and you're managing all of the different batches, you have all

18:54 these different densities and viscosities like horizontally along the line, plus all of your elevation changes. So you're watching all of this in the line of this delay, which is like this long

19:02 list of numbers, essentially your pressure, your suction pressures, discharge pressures flows, and you're watching all of it, and you're manipulating your suction pressure subpoints or discharge

19:11 depending how you operate, to maximize that flow rate all time. So, or to get that thing started up, whatever you're doing. The difference with crux is now you go and you select I want flow path

19:20 A, you know, from A to B, and I wanna hit max rate, and you click go.

19:27 That's it.

19:30 So it's like, it's just a completely different experience. Okay, so let's talk about pipe bot then. So like what's going on? I mean, is there some kind of like AI, ML? It's ML, it's a closely

19:40 automation. So it's model predictive control, which as I said is heavily commonly used in chemicals refining pulp and paper pharmaceuticals. So we've taken that model predictive control algorithm

19:52 and we've applied it to pipelines. We're the first in the world to do it. We actually have a patent on the application of process patent. I don't know why, no one's ever done it before Exxon was

20:00 the only one who had anything even close in the prior art. Yeah.

20:07 And then I'm assuming is Gather bought a similar implementation but just from a different side of the process. Sure. Yeah, so we connect by OPC. It is a read-write. We're currently on-prem, but

20:20 working with - because clients want to see more uptime, right? And our clients are starting to understand what we were talking about earlier about that zero trust security. So we are partnered with

20:28 Microsoft and Microsoft invested in us. to enable that zero trust environment and more connectivity for their applications. Yeah, so that's my question, is where do you live physically from a

20:40 software perspective? Are you deploying on their prem? Are you cloud-enabled or both? We deploy on their prem, and then the plan, we're building this out now, we're not there yet, the but plan

20:50 is their hybrid cloud is where we deploy the cloud functionality. We are able to get into it, right, and access it, but it's all theirs All right,

21:01 how does it work? So, you mentioned the Purdue model earlier, and I've had to work with that or around it, or however you want to talk about it. 'Cause I'm the data guy, I'm like, give me the

21:09 data, you know, and I was trying to tell it. Yeah, you're trying to get around it. Give me one little pinhole into your, you know, fortress. Give me the sense of data, please. You know, so

21:17 I can pull it out. It's totally fine, it's data in, right? So like, like, end device data out, like, I don't see any reason why they can't be giving, stuff like that, especially if there's a

21:26 lot of value

21:28 from like maintenance or things, right? Well, they were able to give data out, but with the Purdue model, they can only push out. You can't, like, I was like, can I, well, I was like, can

21:36 I even request? Can I have my little thing that goes in and requests the data and pulls it out through this one little pinhole that no one else will know exists? And, you know, we can secure it

21:44 with SSH or VPN or whatever the hell you wanna do. And it's like, nope, like, so we had to work with, we had a consulting company or whatever that did a lot of software work for us that was

21:54 pushing the data out for us But we had to pay for the service. I'm like, if you could just give me that one little pinhole, I could get the data out, you know, and then it's snowflake and get it

22:01 to our people. But, you know, I think probably, I think ultimately with that model, you have to be inside of their, you know, DMZ, essentially, you know, like, to be a mentor. To control,

22:11 we are inside the DMZ. We're inside the control system, right? And it was, you know, to the point of like, we've been at this a long time. Like, yes, we have from like a software company

22:24 standpoint But if you think of it, like. we are sending commands that are manipulating multi-billion dollar assets. Like, that's gonna take a while for folks to trust, even despite how big the

22:35 price tag is, and our industry is conservative. And so, yes, it took us a while to get there, but like, I'm extremely proud of the fact that we've gotten clients to do this. Like, they are

22:43 doing it, and every day. Now, that's a very understated, or underappreciated thing right there, is that the whole change management within big energy organizations is a monster, especially when

22:57 it comes to physical assets and controlling those assets, because that's all they think about is risk, right? Like, people don't understand that when they first get into the industry, is like

23:09 everything boils down to risk at pretty much every level. And so, the more things are rotating, or the more pressure there is, the bigger the risk exponentially, essentially And so that's, yeah,

23:23 it's. I mean, even getting them to go to the cloud is a whole another can of worms that is especially with this type of stuff, coming from the historic way of doing things. That's kind of one of

23:36 my questions is like, kind of describe the historic architecture of how these are architected and then describe what you guys see as the future of this for any kind of modern company to really start

23:50 excelling where they're going to have to go Like on the OT side. Yeah, like I said, I worked, so before I started it, digital walk-outers, I worked for an edge hard work startup and so I got

24:03 super into the weeds on the manufacturing and oiling asset on edge and those sort of things and it's an interesting concept just because it's like, well, you spent the last decade selling us on the

24:15 cloud and now you want us to go back to on prem. But in the manufacturing and control system spaces, it does make a ton of sense for a lot of different things. So I'm curious, what's your thoughts

24:30 there? Yeah, so I think like in the, 'cause the manufacturing spaces are gonna be different than wells, like

24:38 a field of oil wells or long haul pipelines or even refineries and stuff, manufacturing seems to be better contained, if that makes sense So in those situations, pardon me, it doesn't move, yeah.

24:50 And so they're like this straight to the end device type control systems might make more sense. Like there it could actually make sense 'cause you have like a contained environment like overall,

25:02 where as like in these more distributed environments and you have like custody transfers and stuff, you do want like, you still want those gates of safety that the Purdue model provides. You just

25:11 still, you want to maintain those while also creating conductivity, right? It's like you have to do both at the same time So I don't know if I fully answered your question if I tangented that this.

25:20 Well, what's the, like, what's the kind of historic architecture for the IT OT side with, you know, there is no connectivity. So do is it, right? Which is like, which is everything's

25:30 airgapped, essentially like imagine in the 90s and you were installing stuff on a server with a floppy disk. That's how it's designed. That's how this stuff is designed. And so, you know, the

25:38 companies that that own this software and that have built this software, they built it 34 years ago, and they built it fantastically 34 years ago. It's just like any company you're big and you're

25:47 successful and everything's working so you don't, they had no reason to rip up what they were doing, right? So, now we're at this interesting point where it's getting, yeah, it's getting

25:56 complicated, but yeah. So, before everything's on-prem, there's absolutely no connectivity. Where corrects is going is there is connectivity, but only within the private cloud that the client

26:07 has. So, yes, the client has cloud. Like, when people hear cloud, they think public cloud, but cloud's just a technology, right? And so, Microsoft can bring in a cloud for a client. right?

26:19 And they can have all of that redundancy, all of that backup, all of those wonderful things that they're used to building for with server banks in their skate environment. They can have all that in

26:27 cloud. So that's the new, that's the new, I guess, architecture that we're seeing that folks really want to go to. And we do have clients that have actually signed up to go to it is like, okay,

26:36 everything is still very contained in your DMZ on-prem. That's where the executables actually run. However, there is connectivity out with the proper gates and the proper, like, everything is

26:47 locked down all the time unless they want it to come in. Yeah. That's what I call what we had at Grayson Mill. I mean, like, we had, you know, it was kind of on-prem, but then we had like the

26:57 Azure, you know, for any kind of backups. Yeah. Yeah. Okay. So you mentioned earlier that, say, Exxon was probably, so many saw that had anything even close to that. And then, you know,

27:08 I'm saying that you worked at TransCanada or TC energy. I think we were talking about this before we jumped on, but I, I done like a really small like consulting engagement through a a consulting

27:18 company with them and they've been working this auto pipe kind of thing where it was just trying to create recommendations ML generated recommendations to the operators to help them optimize the flow

27:28 of the pipe. I mean they were trying to get there and so was any of that trying to go on when you were there and was that part of this you know jump that you took? I think it started because of me

27:38 because I was the first one pitching that I'll just take the credit because I was pitching it there three to four years of right and left and I was in the control room all the time. So I think it was

27:51 I left because I wasn't able to get any traction which is like totally make sense like big companies are not built to like try new great stuff like they did the right thing as an organization. I

28:00 loved it there it was a great time. I found it really interesting that they were trying to do it in-house after I left. Yeah I had multiple folks call me asking me for help with it, actually people

28:11 who knew me. And I was just like, no,

28:14 I'm not going to be an early doctor of Crocs, but yeah. Yeah. I was like, you guys can have the Crocs number and we're happy to do it right.

28:22 Make it work since, especially with Keystone. Keystone's biggest customer is Phillips 66 and we're all over Phillips 66. Beautiful. Um, so as far as like, you know, like, and maybe this is

28:36 more of a question for those CTO type, but I'm sure you at least, you know, no high level I mean, I, what are these applications written in, you know, as far as like, like a technology stack

28:45 and, or like, how are they deployed? I mean, is it containerized or is it just Python, yeah, fully containerized. Yeah. Which is like that's, I learned about that. Yeah. Cause it's like,

28:57 okay, the difference between something that's like floppy disk installable versus something that is really flexible and can go anywhere. That's containerized. And, uh yeah,. So we're, we're

29:07 fully containerized. The client is like, hey, we don't want this hybrid cloud thing. We really only want you on prem. We can - sure, we can absolutely do that. And when you're ready, we can

29:15 then enable the hybrid private cloud thing if they want it. So we're - because we've designed with a modern architecture and modern capabilities, we are infinitely more flexible as a package than

29:28 the existing stuff.

29:30 Walk me through - so I'm a client. Walk me through how do you need - test that - how much data do you need historically? Three to six months historical data is what we

29:37 like to have. So

29:45 we are first principles physics-based at the beginning and then augmented with machine learning. That's extremely important because you don't need a bazooka to solve a problem that a hammer can fix.

29:59 And people really miss that point. So model pretty to control is first principles based. We know Bernoulli's equation we can get most of the way there, right? So that's why you only need three to

30:10 six months of data is because you first principles know what this thing's supposed to do, right? So we get that data, then we have a simulator of the asset. So like a digital twin, that's like a

30:21 transient hydraulic simulator, so it mimics the dynamics of the actual system, a lot of people do steady state, that's stupid, you're only getting point solutions. So we have a transient dynamic

30:30 simulator, we test our software on that transient dynamic simulator for a series of operations that a controller operator would totally, would typically be executing on And then we roll into

30:39 production with really heavy buffers, so making sure that we're well away from any safety limits, roll out in production, get control and operators used to it, and then we slowly open up the

30:47 buffers and maximize the throughput. Control and operators were seeing about 83 utilization, kind of on average, like they love it, like they ask for it when it's down. And so when you first

30:58 start out, is it kind of like human in the loop feedback where it's just making recommendations, or is it full control, full automation? Full control, okay, so I'm gonna get a little spicy The

31:08 whole recommender thing in pipelines, I think, is absolutely BS, like 100 BS. And the reason why is why would you spend millions of dollars to build something that you can never measure the ROI on?

31:19 You cannot tell if that control room operator listened to that recommendation and actually did that thing or not. So therefore, you have zero value without closing the loop. So in my mind, it's

31:27 like, okay, well, we've already closed the loop with this exact technology in every other industry in the world. Why are we so afraid? Let's just do it. It works. And so we did it You can take

31:36 three months of historical data before we implemented crux. You can take three months of historical data after. And there is, you know, that two to seven percent, 300 barrels an hour-ish increase

31:46 that is a measurable result that you can then sell, like contracted volume.

31:52 Yeah, no, that's, well, especially, you know, that's, that's one of the things that people just don't think about from a data perspective, especially when it comes to any kind of industrial

32:02 system or controls type stuff is that the human fatigue element, like, We are very fallible and we get tired or we have drama going on our lives or we like flipping through Instagram like humans are

32:16 not good at doing these. Yeah, and so those are things that we should be automating because computers don't have emotions. Ninety five percent of heavy industrial accidents are human error. So

32:27 crux reduces human human intervention by 80 to 90 percent. That's an 80 to 90 percent reduction in potential human error, right? Like it's like it makes sense. It's what you're supposed to do It's

32:37 what we do in planes. It's what we're doing in cars. Why aren't we doing it in pipeline? Yeah. And that's one thing that's always been, you know, the thing being in the kind of upstream side,

32:46 especially I've worked a lot with reservoir engineers, like theirs is kind of a finger in the air kind of engineering and, you know, there's a lot of uncertainty and ranges, but like when you

32:54 start getting into like a pipeline or even facilities, like I know the diameter of the pipe. I know the properties of the pipe. I know the properties of the fluid. So like again, like these are

33:02 all. I know friction factor. I know everything, right? Yeah. So I mean, like computers are excellent at that.

33:08 Fabulous at it right and like and I we even have a guy on stuff like a PhD in numerical methods and like That's how you get the rest of the clothes on these types of solutions right like you know

33:19 physics can only get you so far But it's a way better starting point because you don't need mountains of historical live feed data or anything like that You don't need to train your models in the same

33:28 way, right? Yeah

33:35 So, as far as, I mean,

33:39 if people have a hybrid cloud, so obviously, you know, you're partnered with Microsoft, so I'm sure Azure is, and probably, you know, in most cases, probably most people are using Azure. But,

33:49 like, if someone's using AWS GCP, because he built it in such a way with, like, you know, containerized things, like, they can use that if that's what they've got. Yep, yep, exactly So,

34:01 okay, so you brought up, like, fabric earlier. And so, with fabric, I mean, like, that's, well, Microsoft is a marketing powerhouse, savage, also.

34:15 But, because, I mean, now I open up Power BI and it says fabric or whatever, you know, but, like, so, now how, what does this look downstream of, you know, crux, I mean, like, so crux is

34:24 enabling, like, you know, are the things that people are able to get out of crux that they're able to, then, you know, extra metadata, things that they didn't have before that they can use.

35:28 so that they can see the operation. They don't need access to SCADA. Like you can't even barely understand SCADA. Like they need like a dashboard that just tells them what like you're up and

35:35 running and you're making money. Yeah.

35:40 For sure. So you see that people are able to connect. So are they connecting directly to the crux OCM like kind of like software or is there a database under the hood? Crux to fabric and then

35:48 fabric would would disseminate across the org. Okay. So when we're saying crux to fabric. So fabric ends up being, 'cause I mean fabric is kind of also an answer 'cause they databricks or

35:57 snowflakes. I mean, like, I'm assuming that's writing to some, you know, like data lake, data lake house type. Yeah, to like whatever data like they have. And then like, then yeah, then

36:05 Power BI would tie into that for whatever dashboarding people really wanted. Gotcha. Are you all able to optimize? Because you're optimizing controls, but I assume you're also able to optimize

36:17 maintenance and things like that as well, right? We're not looking at that right now because, you know, software company can't boil the ocean. It's like you could do anything, what's the highest

36:27 ROI for the client that's like the biggest evolution. Yeah. Yeah. More barrels. And like, and most of the incumbents, like the existing big OT folks, they've got, they actually have fantastic

36:38 maintenance solutions. Like, like that seems to be the, the angle that they've been able to tackle it and then get better data. And like, so it's just not somewhere where as an early stage

36:46 company, we, we would want to compete. Like, I think those guys are way ahead

36:55 of us Yeah, no, that

36:60 makes sense. You're, you're right. I feel like everyone's like first stab at, you know, doing something with their industrial data is, is definitely maintenance. Yeah. Yeah, we probably do

37:02 some things better because we analyze the control loops. The control loops are typically like you're, you're indicative of anything that's going to go wrong because that's your actual operation.

37:12 And so there's a possibility on the road that we could look at maybe doing it better. But like, again, from the stack of priorities the clients would want, that's probably a lower one. Yeah.

37:23 Okay, so now I'm bouncing. back to the whole other side of the stack. So tying into like SCADA. So I mean, I'm assuming you're able to tie into a SIGNET ignition, you know, any of the major,

37:35 you know, and how was that, you know, early on, trying to like, did you focus on one of those first and then tackle the other? Or was it just kind of like whatever you got? It's all OPCUADA,

37:43 so, or an QTT. So there's really only the three protocols and we can handle all three. So it's fine. Okay, so, and that's maybe something I didn't necessarily understand. So like most of those,

37:54 you know, whether it's significant or ignorant over like, they speak a common language out. Yeah, OPC. So it's like open process something. I call it O-Please Connect 'cause it ain't the best,

38:03 but.

38:05 Touche.

38:09 Now, that's another question I've got is like, you know, Bobby and I, when we worked together, we were working at kind of a sensor company. And so we went into MQTT versus, you know, the

38:22 PubSub and all the different kind of methodologies there. What is your kind of,

38:29 I guess, take or stance on, on which, I guess, protocol is people should be using or at least moving to it, right? Like there's ModBose, there's TCPIP, there's all these different connection

38:42 and communication protocols and stuff out there and it's like, hey, if we can just standardize this, this would make everyone look like a lot easier. I agree with you on that. I think

38:50 standardizing it would make everyone like easier. I don't like, yeah, one of our other guys, like Dale, especially, like he's a former Phillips 66A. He's our head of deployments. Like he would

38:58 have, I'm sure he would have a lot more on like, which of those he likes on the deployment side. From my perspective of here, I'm kind of like, as long as it connects, we're making money, right?

39:10 So that's what we want to be accomplishing.

39:14 All right, the question that everyone always wants to ask and I'm gonna make sure it comes up. And here is like, is our LLMs or AI, how is that impacting business or is it being used? I mean, I

39:27 know we've talked a lot about on the podcast that it's not necessarily ready for prime time, we'll say time series type data, but just curious what you guys are seeing. Yeah, no, it's not quite,

39:36 but what we're doing with it, which I'm really excited about, is with advanced process control technologies, people's biggest complaints is like, well, what's it happening in the black box? We

39:48 don't know. So we're using LLMs and we've got like our first demo set up and because we are containerized and we have the connectivity, we can actually use them where you can ask the system what it

39:59 did. So instead of a recommender and the system telling you what to do, the system's doing its thing. And then if you're like, okay, why did it do that? Like, you can ask it why it did that and

40:07 it'll tell you, right? You can ask it to open a ticket. You can ask it to help you investigate issues, right? So that's what we're working on LLMs for because if you think of, like yes, you

40:17 wanna be able to query your time series historical data, but maybe that's more of like a formulated dashboard. But in terms of what the application did and why, like that's a training, that's

40:28 troubleshooting, like that's hugely valuable for new control arm operators, right? Yeah. So that's where we're going with it And, you know, to the point that we're talking about before, of like

40:36 all of the existing OT software not being built with connectivity in mind, like they may be able to show a flashy LLM demo, but they actually cannot, like they cannot generate the connectivity that

40:46 makes an application like that work. It's actually physically impossible.

40:51 Hard to have a LLM if you don't have any data to feed into it, right? Yeah. No, that's so -

40:59 So - That's honestly good, Bobby No, I was gonna say with that.

41:05 You know, so I think there's two sides of this, right? Like, I mean, now that crux is essentially almost doing these people's jobs for them in a way, you're right. I mean, like - Helping.

41:13 Helping, no, obviously helping.

41:16 But I mean, they don't have to do as much as hands on as they used to be. So like, obviously then this LLM type thing can bridge that gap and can help train people that don't have, but like - Keep

41:26 them super sharp and engaged, right? Yeah. Like with it, 'cause it's an interactive system. And like, what we find too is, yeah, like folks really like it. And like, what they do, like the

41:35 way that it works just to make that clear too, is like, it's like, if you're on cruise control and your car and put your foot on the brake, so if they touch a set point, the whole system kicks

41:42 out and they're back in control. So we really like that from like a control or operator engagement standpoint. Like we very much think of it as like, okay, pilots and planes, right? Like you

41:51 need a pilot still, but you also really need autopilot software. Like that's not safe in a commercial plane without one, right? So we think of ourselves like that. Okay.

42:02 I mean, this is an interesting thing, and I think we've talked about this with AI or just automation in general, but like, because I think I think it's big, probably just from covering your own

42:10 bases, like, because otherwise if you were pitching this thing that this will just do it for you. And then something goes, hey, why I mean then liability probably would shift to you where like,

42:18 no, I mean, like, this is still in the control of the operator, you know, it's just making, making them better at what they do, but they can take over whenever, you know, so that burden is

42:29 not back on you all to be perfect To be perfect all the time. And that's the thing when people go straight to the end device, like, your, your bar for perfection is so much higher, right?

42:40 Whereas, like, trucks, we can install in production. And, you know, heaven forbid we send a wrong set point. Well, Skate is going to stop it and the PLC is going to stop it. Right? Like,

42:51 it's just not, it just can't get through. So, like, it's just so much safer from a control standpoint to, to structure that way. Yeah. I think the autopilot example, that gives me the perfect

43:03 understanding of what it is, what its intention is. It's not full automation, it's just there to autopilot and then anyone can control it when they need to, right? Well, because Keystone,

43:14 starting Keystone, I'd break a sweat. I'm like, this is nuts. Yeah, that's wild. I had no idea. This is wild. I should not be breaking a sweat in a day job and when everyone's getting all

43:25 pumped about it, I'm just like, no, this isn't seen, why is this a job? You guys are not getting paid enough for this.

43:32 Yeah, especially if you can code it, just codify it and it says that's the, when you're doing these repeatable processes and humans, especially when they're long and focus intensive, humans are

43:45 not good at those things and they should absolutely be automated. Yeah And, you know, folks are amazing at their jobs and, you know, we don't, when we come in, like we're super conscious of

43:56 that, like we don't want control room operators to feel This is there to replace them because it's really, really not. Like we built it, I built it because I was breaking a sweat and I was like,

44:06 Your guys' job is too hard. And then we realized like, Oh, there's a massive ROI to be had by doing this better. And that makes sense. 'Cause if you think of it from a first principle standpoint,

44:14 right, your control room operator is literally touching your PL. So if you make their job better and easier, they should be making you more money. Yeah.

44:23 Yeah, help me help you win. We all win, all right? We all win, yeah So I guess with that, how about, I mean, how much has this impacted companies say from say an ESG standpoint? I mean, like

44:36 have they been able to, obviously people, if they can allow more throughput, especially to say on the gas side, are they then able to, you know, they're able to flare less or, you know, do

44:46 certain things like that? Like have you seen like a tangible ROI on that or has it come up? Yeah, we have. So we, when we were first in production, we had forgot about a functionality that

44:58 control room operators like to do, which is called holding a leaning profile. So it's where they pull the suction pressures at each pump station to as low of a set point as possible hydraulically.

45:09 And that's from a safety perspective, right? 'Cause they wanna operate the pipeline just as far away from NIOP as they can normally. So when we first commissioned pipe bot, it didn't really have

45:17 that instructions and it was hovering a little bit too high. So we put that functionality in and pulled it down. And what we realized is that then at the same time that you're maximizing throughput,

45:27 you're actually losing, you're actually using up to 10 less energy per barrel.

45:32 So it's, yeah. So that's already fully commercialized in pipe bot. And then we are looking to commercialize now with the emissions reduction Alberta power out, which would take power contracts

45:42 into, into account. And then be able to select which pumps should be operating in which geographic region to minimize costs. So that's something we're working on as well. But right now, yeah,

45:52 the ESG benefit is, It's real, it's like, okay, that's interesting. Right now, our clients aren't reporting on it, but we are measuring it, so whenever they want it, it's there.

46:04 So you're talking about going all the way to the supply, like connecting all the way to the supply side to be able to afford effectively optimize for real-time demand. Yeah, you just wanna have

46:14 like the power track contracts as an input, and then you wanna know like, okay, if this is your oil schedule, like your nomination, what you need to move, then if you can forecast it out like,

46:25 'cause oil nominations are usually a month, you could forecast it out two weeks, and then what happens is you end up with like almost like a daily schedule of how you would operate your pumps, and

46:33 instead of like, like they do this right now with engineering, right, and they'll send the instructions to the control room, but again, if you can't control it, you can't optimize it, so you

46:42 can run all your simulations, all you want, and send the control room all you want, but like, you have no control of whether or not that gets executed. So what a power-up does is actually execute

46:51 that, to run the simulation execute. That's cool.

46:56 Yeah, when you get even more of that context and you understand like all the external factors that are impacting it,

47:04 you can plan for it. What about weather? I mean, how does weather, are you pulling in? It's minimal. They pull any weather. Okay. It's minimal because it's like, because it's like, you know,

47:13 in the, like, you might wanna move a little less in the summer, but like, depending on temperatures, if you're having some measurement quality issues that you're concerned about with vapor

47:22 pressures or something, but again, that would come from, like, your measurement engineering team saying, like, okay, we need to slow down a little. Like, that's like, and that kind of thing

47:30 would happen like a few times a year that we've only seen. Like, it isn't like a, yeah, it's not like a thing that you'd wanna be able to automate for. That's the same thing, like, we get asked

47:38 about pigs, and they're like, whoa, can you automate, like, all of your pump station shutdowns for pigs? And I'm like, well, we could, but you guys piggy each line, like, once a year, like,

47:46 why would we? Right. It's prioritized. That's not gonna make you money, client.

47:55 And then from your side, John, that. No, this is fascinating. I'm happy to see someone optimizing our energy infrastructure. Intertracting. 'Cause it is, it's kind of terrifying to think about,

48:08 honestly, when you think about how old it is and how much, how just archaic most of it really is. And how many, and how like all the folks that are like, like, you know, blessed them, all

48:17 these are all great companies, but they're all like just antiquated and like the leadership in these companies don't even realize how antiquated what their products are. One there, but they're also

48:24 caught in that conundrum of doing more with less, spend less money, you know, all of this stuff. Oh, but also invest in technology, which costs money and takes time and change management. It's

48:35 a tricky place to be, but - Super tricky. That's why you want to talk about it on the podcast. So. When it's good to get someone, you know, a midstream, we haven't had a ton on the midstream

48:46 side, or really midstream or downstream, but like I think most people think of oil and gas and they think of upstream and they think of - downstream and they don't think about the mid-stream until

48:54 well they forget that it's all connected so like everyone's so hyper focused on optimizing refineries optimizing wells and it's like All of that is connected by pipelines people and if we're all

49:03 operating that completely manually that 26 miles of pipe in the US alone Like it doesn't matter how much you optimize your wells or your or your refineries Yeah, if you can't get it to the refinery,

49:12 you're not there. No one's making money Yeah, until colonial pipeline happens and then you're like oh wait Why is my you know gasoline going up, you know, it doesn't make any sense. Yeah, like

49:23 wait, all right

49:25 Important commodities Yeah, no hundred percent well that and people don't even understand well we could get into the whole thing of people don't even know The difference between natural gas and

49:33 gasoline You know

49:36 half the time, but I mean, how does natural gas get to your house? Just it's basically through it. Yeah, my kid asked me that the other day. I was like, oh man. That's a long conversation I'm

49:47 going to tell you every single part of that. You get the whiteboard out and you're like, No, yeah, this will be a multi-day lesson here. But keep asking. It's important. Absolutely. That's

49:60 super important. It's, yeah, it's, yeah, the, just didn't, you know, we talk about it way too much. I don't know, I don't know, I don't know. So we're like just the general population's

50:07 lack of understanding of the energy sector and how it powers their entire life is almost comical. Yeah. Yeah. Well, I guess even now, especially with like the upstream and midstream kind of

50:19 joining forces, but also just with AI and our data centers, I mean, like, what are you seeing because you're working with people that are doing this, I mean, like feeding natural gas into data

50:27 centers. I mean, is there anything around that as an impact on what you're doing or? There's definitely like, I definitely heard conversations like at zero a week, like you get these little

50:36 snippets of like, oh, yeah, like Microsoft and then this energy company are looking at this data center. And like, so it's definitely like, like tech in oil and gas are doing way more of this

50:47 then I think folks. realize.

50:53 And I think folks thought we were already doing this, but we weren't, right? We weren't quite there. Yeah. Well, you're out in Silicon Valley. I mean, like, what are you hearing from people?

51:01 And I know Collins been out there a few times too, but like, what are you hearing from the kind of boots on the ground of people, like their thoughts or feelings towards say, oil and gas or just

51:12 energy in general? I mean, I've been saying here now the last couple years, like her last six months, especially like, if it takes Silicon Valley being the one that pushes us to build more

51:22 nuclear and more pipelines and more to get more energy than I'm here for it. But seemingly for a while, we were almost kind of demonized in some ways by people and that's out of the way. So yeah,

51:36 so this is curious what that's like for you being the person from the energy side, but also straddling that fence. Right in the world. Yeah, that's an excellent question and I definitely think

51:46 it's changed.

51:48 three years now and had spent quite a bit of time before that because a lot of our investors are here and it's definitely changing like in terms of the Silicon Valley talent when I first tried to get

51:59 like executives here, a lot of skepticism. Now like we're currently doing a search for like a finance executive and I've gotten this feedback from a few folks that we've gotten recently, it's like

52:10 they're looking to work on things that are more meaningful, like that's actually the sentiment. They're tired of optimizing ads, right? So you're, it's really cool actually, you're getting these

52:20 really smart people who are like, okay, cool, like I optimize some ads for a bunch of people in the Midwest, but like who cares? And so we're getting talent now that like have a lot more time for

52:29 companies like us, like they're not just like, ew, oily gas, they're more like, oh, energy. And then also too, so it's definitely changing, which is fantastic. I think the overall sentiment

52:38 is much more positive, like for example, I was fortunate enough to have a round table with the CEO of Microsoft and literally straight out of his mouth is energy is existential. And like, we're

52:49 the first energy focused company like for software that Microsoft's M12 has invested in. So like, they get it. So, you know, the general population of California may be, of course, lagging,

53:02 but like, I think the big, the big hunchos in oil and gas are definitely not, or in tech are definitely not demonizing us anymore.

53:10 Yeah, I think everyone's quickly realizing that if we're gonna have more data centers, we need more energy And if we're gonna improve our energy operations, we're gonna need more data centers. So

53:22 it's this weird marriage we're stuck in. Yep, yep. But it's good. It's progressing both sides, hopefully. Yeah, it's honestly getting way better. Oh, sorry, gonna bother you? No, no,

53:35 you're good. No, I'm just saying, some people start saying like it's not an energy transition, actually it's an energy addition. Like, and really actually in no point in history have we ever

53:44 transitioned from one field to another? started adding more on, adding, you know, like. Yeah, and sometimes we'll deprecate something, but like. Right, there's always more additions. Yeah.

53:58 That's awesome, man. Absolutely. This blew by Vicky, this was a fun conversation. Oh, I hope so. Okay, did I touch on everything you guys we're hoping to get? Yeah, absolutely. Except for

54:09 the final, final, we need like an audio drop here, Jacob, like a final countdown audio clip right here, but our speed round at the end, it's super intense. We just ask you random questions

54:24 about random stuff, but

54:28 I'll jump in favorite or highly recommended book or books for a founder. My two favorite business books are warfighting, which is the Marine Corps, US. Marine Corps Doctrine and team of teams. So

54:45 you guys do military real good. I like reading about it. So, and I think it's extremely transferable to business. So yes, those would be my two favorite business books. Then also the hard things

54:56 about hard things by then Horowitz. So if anybody, especially in oil and gas, if you want to get like an insight into the pain of building a software company, read the hard thing about hard things.

55:06 Yeah, that was a great book.

55:10 All right, what's your favorite Canadian snack? Oh,

55:18 so I'm from Newfoundland. So my favorite Canadian snack would be jam jams, which are these weird little like jam cookies that are from the East coast of Canada. I also really like poutine. Oh yeah.

55:32 Yeah. Love some poutine. I don't know if either of you watched Top Chef, but they're in Canada this season. And so one of the challenges recently was poutine and I was like, oh man, like it was

55:42 like nine o'clock at night, I'm laying in bed. time to just destroy.

55:50 Speaking of Newfoundland before I forget, do you know Josh grows from pod two? I don't know, but we go back to be related. No, yeah. He's their CTO, but he actually lives in Newfoundland and

56:02 basically develop their application, but really good dude. But, you know, another new fee in the oil and gas industry tech side, so I would definitely recommend reaching out to him. Yeah,

56:12 actually, I'd love an intro. That'd be great, because it's always nice to us from the same little island we tend to connect Yeah.

56:23 I don't remember whose question we're on. Are we online? What is, what's your go to, like, what would you recommend? Where would you recommend someone go eat in Calgary if they were visiting?

56:34 Oh, Calgary, okay, I was just there. I would recommend for a full experience, go to Caesar's Steakhouse. Because it's like

56:45 the OG Steakhouse in Calgary. It's like, like, you know, everything's red velvet, gold, and just like, there's like a fire pit in the middle And it's just like this OG, like, experience. The

56:54 Steak's pretty good. Like, it ain't, like, epic. But it's also, I think, the home of the Caesar, where Calgary invented our version of the Bloody Mary. So it's worth it. Oh, cool. It's

57:03 really worth it, is it? Is it called a Caesar? It's called a Caesar. I didn't know that. Cool, I gotta look that up. I'm a big glimmer. It's got tomato juice in it instead of tomato. Okay.

57:11 So it's just got a little bit of an extra zing to it.

57:16 Awesome. One more thing. What we

57:21 got I'll see your favorite out. Why you got one then? Okay. Yeah. What's your favorite vacation destination? Right now I would say Little Corn Island in Nicaragua because it's just like there's

57:33 no roads and there's no cars and it's just nothing and I could just like go snorkeling and I like it. Sweet. Just totally off the grid-ish almost. Yes. That is. That was my favorite part about

57:44 where we went in the Bahamas. There was one road and it was just roundabouts and I will say going going on a roundabout in the wrong side of the car on the wrong side of the road is a complete

57:56 mind-bender but

57:59 I love to disconnect. My last wrap-up question is what is your favorite Canadian jargon or slang term? Yes, by us Americans have no idea what you're saying. Yes, by or what are you at? These are

58:14 new fee. These aren't actually Canadian. Perfect So they were like extra. So yes, I is like, oh, yeah. Okay. Yes. Bye. And then what he had is like, what are you doing? What are you up to?

58:25 Yeah. I love that. I love that we speak the same language, but we don't speak the same language at all.

58:32 It's legit. Yeah, that's awesome. Well, cool, Vicki, thank you so much for joining us. Where can people find you guys if they wanna reach out or get in touch with you guys? Yeah, please come

58:43 find us on LinkedIn. Yeah, that's where we play lots, but also Vicki, the ICKI at cruxocmcom If anyone wants to reach out directly, we are always looking for more customers who are willing to

58:54 give this stuff a go. Awesome, that's

58:59 CRUX. OCMcom. Perfect. Thank you.

59:03 Awesome. Well, Bob, do the influencer thing and then take us out. Yeah, make sure you like or subscribe or follow us or whatever the hell

59:14 you do on your preferred platform. And we just appreciate you guys watching and keep it up See y'all next time.