Chuck Yates Needs A Job

0:20 All right, so how did you even get on my calendar? Literally, I think we were supposed to do this Friday, right? Yeah, tomorrow. To do something? Yeah. I had this on my calendar, I get an

0:32 email that says, Hey, can we do it again? How did that happen? I mean, our marketing team, I mean, they are looking for a great podcast and great ways to send the message. So they put it on my

0:45 calendar, and then I just found out, and then I realized that, oops, you know, I have another comment on Friday, so I asked whether we can move to Thursday. Thanks for doing that. Do you have

0:55 kids voting chance? Yeah, I do. Well, you know how kids are always like, Hey mom, dad said it's all right. You know, Dad, mom said it's all right. If you said it was all right, I think we

1:06 got done that. Yeah. Nice to meet you. You too. So,

1:14 I was looking at kind of a company website, trying to read through what you guys are doing. tell me about your company. Yeah, sure. First of all, thanks for having us here, having me here.

1:25 Uptime AI is a company and what we do is we improve, we help companies or we improve the profitability of manufacturing companies in specifically in heavy asset industries like oil and gas,

1:39 chemicals, LNG, SIMM and power generation, renewables, etc.

1:45 What we offer as a company is a software solution that enables operations teams to run the plants and solve problems like experts. I'll give you

1:58 a quick example. For example, in

2:03 a refinery, there is a problem. Let's say the motor is getting hot.

2:10 What do you do? You go to someone or a senior expert to try and understand why it's getting hot. What do I need to do? Right? And the senior what they do is they then say, okay, what is the

2:21 motor connected to? It's connected to a pump, let's say, right? And what's a pump doing? It's pumping fluid. And now what's happening with the fluid? The fluid is getting thicker. Oh, that's

2:33 why the pump is overloaded, and that's why the motor is getting hot. You see, there is a there's a process of actually looking at the symptom of a problem to actually understanding why what's

2:44 causing it and how to fix it, right? Today, most of that is done by experts, right? However, in this industry, we find that that kind of expertise is going away, and the younger engineers,

2:58 you know, we don't find people who want to stay in the plan for 30 years to become those experts. So what do you do when you are losing that expertise? And that's the problem that uptime AI is

3:09 solving. So our solution acts like a virtual expert, which is connecting the dots between your equipment. different types of equipment, different functions like the electrical equipment,

3:21 mechanical, and the process to kind of, you know, solve, connect the dots, solve problems, and learn on the fly like an expert. And the beauty is, you know, that's why we call it as an

3:34 operations excellence solution, because it's not looking at just rotating, it's not just looking at the process, it's looking at holistically the entire operations The sun's dangerously close to AI.

3:46 Are we talking AI? Yeah, exactly. So the software name is AI expert. So it's it's it's acting and mimicking how an expert connects, dot solves problems and learns on the fly. So I want to

4:01 actually go wonky and geeky on this and just a little bit. But my mom watches the podcast. So and she's really sharp. I mean, she went to Rice University, smart on smart cookie. Shouldn't know

4:16 anything about AI. Can we use? 'Cause we hear about AI all the time now these days. Right, right. Simplify it down. Okay. Tell me what that is and why are we talking about it now and we weren't

4:29 talking about it 20 years ago. Yeah, great question. I mean, there is, honestly, there is a lot of material and marketing, you know, stuff out there to the extent that it's been in some ways,

4:42 you know, bastardized, you know? Like, so the way we interpret AI in the company and we use this definition across, you know, all functions in the company is can it mimic the way people are

4:57 solving the problems and learning, right? If an application can do those two things, like solve issues the way that a human can do and, you know, learn on the fly, that's what, you know, we

5:09 call it as artificial intelligence, you know, based applications So, if. If you see our solution, what we offer, as I said, those two elements are key. A human understands the context, and

5:25 the context is basically what differentiates between how a human can understand different pieces of elements and solve the problem. And what we, for our application, it is able to connect the dots

5:41 between your mechanical electrical or your process and reliability. That's how a human would do. That's one part. The second part is, can it learn on the fly

6:00 on its own? It's not repeat, it's not automation, you know? A lot of people confuse automation with AI, but it's not automation. It's about figuring out how to solve the problem on its own on

6:06 the fly, right? Now, let me cut you off just to drill down on that, 'cause I think you're right, you hear that all the time So automation might be. Rules are imposed upon the machine, and the

6:18 machine does it in effect automatically. So turn off the lights at eight o'clock. If the temperature hits this, shut it off. That's automation, but it's not thinking. Exactly. Okay, I got you.

6:31 So if you tell a rover to go from here pointed to point B and you

6:38 code it up, it will always do that, right? But if you put a chair in between, or a block in between, if it can actually figure out and then move around it and then go, that's actually learning

6:50 on the fly and solving it, right? And that's what we're calling, right? Have you seen that with the Tesla self-driving cars? It's supposedly people are dressing up like stop signs and making the

7:01 car stop. Yeah, I

7:03 mean,

7:06 the space is evolving quite a bit, right? fascinating to see how these type of solutions are growing. Right, and so the reason this is happening now is they're kind of a seminal moment of some

7:22 sort, or is it just the combination of computing power, et cetera, that we've gotten to this point? Yeah, I think - 'Cause it feels like it's exploded. Exactly, yeah. I would say certainly the

7:36 compute power, the data available with the, right, and the evolution of deep learning and techniques, right, and they're used by some of the software companies around the globe, they have

7:48 certainly accelerated the journey towards it, right? But also what I would say is there is a technology and there is a need in the market, right? And when those two are kind of meeting up, that's

8:03 when the explosion really happens, right? So today, if you see most of the AI-based solutions, they are specific to particular problems. You have, you know,

8:16 sales agents, right, which is actually like an AI agent, which is listening to a sales conversation and suggesting what to say, right? Or you have, you know, agents for customer service, you

8:29 know, but there are all for specific type of problems. And they're adding a lot of value in making people more efficient and more effective Gotcha. So before we drill down into what

8:44 you guys are doing, I kind of got two questions. Yeah. One of them will go wonky philosophical on you. But first, is there a way to simply describe how a computer is learning? Yeah. Yeah. I

8:58 mean, because I mean, I get your difference on it, but how does a computer actually do that? See, the best way to explain that is let's take a child, right?

9:11 Basically, the child as they grow up, right? Um, he or she grows up every time they do something, you know, either a parent or someone or, you know, there's a feedback, right? Um, it's

9:23 either a parent or they get hurt or, you know, they get rewarded. Right. Um, and that feedback is essentially forming a pattern in their, in their brain, right? I'm saying that if you do this,

9:34 this is what is going to happen, right? And those patterns are what are forming that knowledge or that learning Right. The same thing happens with a machine. You know, you call it reinforcement

9:48 learning or, you know, um, basically. The machine does something, right. Um, and for example, in our case, the computer or the software is going to tell you, Hey, your compressor is

10:02 following, which is why your bearings are going to vary out. And so you should do this, this, this, this, this, right And the user, the engineer can say, yep, this is a gradient site. I'm

10:13 going to do exactly that. Or they may say, you know, it's not as bad as the software thinks it is, right? I'm not gonna do anything. And every time they're either doing or not doing, the

10:27 application learns that, okay, you know, this is good. I should do it more. This is not good. I should try and optimize my understanding. So that's exactly how it's learning. Oh, wow. And so

10:41 the, 'cause what you hear is the internet, all the conversations, whether that's the repository of Reddit conversations, Twitter, whatever, you have, in effect, teachable situations, right?

10:57 You had interactions. Exactly. So the computers learn how to react, gotta help us that they're training us on, you know, 2024 conversations. But anyway, so

11:12 before we jump in though, I want to address this because. I find this fascinating.

11:20 If you look at the spectrum of opinion on AI, over here, you've got Mark Andreessen. Mark Andreessen, this is a fancy hammer. It's a tool for humanity. It'll make things better. If it gets too

11:34 bad, we'll just unplug it. Right, right. I mean, he's, I don't wanna say he's utopian, but I mean, he defends the nuclear bomb as being technology that has made the world a better place. On

11:47 this side of the spectrum, you have Tucker Carlson, who's literally ready to turn those nuclear bombs onto every data center in America. He views AI as should we be killing baby Hitler, right?

12:02 'Cause it's in its infancy here. Right, where do you shake out on that moral spectrum? Yeah, I mean, I have opinions on the application of AI, right? I think, you know, most of the time.

12:17 the technology per se, right, can be used for a good, can be used for doing bad things. And it's applicable for everything, right? You can use a, you know, you can use, you know, a rifle for,

12:29 you know, just recreation and have great time, but you can also use it to, you know, cause damage, right? My stand is that as a technology, it's a great thing, 'cause it's gonna make life

12:43 better, right? How we use it is something that needs to be understood on a case by case basis and say, you know, and say whether it should be done or it should not be done, right? For example,

12:58 you know, I have opinions on like, I'll give you a short example. Like let's take YouTube shorts, okay? The shorts or the quick videos that come on YouTube. Now, even for an adult, right,

13:11 when you open it, okay?

13:15 I find it hard to actually stop it, you know, in less than five minutes, right? It's crack cocaine. Yeah. Exactly, right? Now, okay, for an adult, it's fine, right? But if you start those

13:24 with, you know, for kids, right? You don't stand a chance to go against the, the intelligence in terms of how it's picking the video that would engage you, right? Because don't stand a chance.

13:39 I mean, you know, and so how much do you, where do you apply? How do you apply it? I think, you know, there are, um, I think there is, there's a balance to be had there. So I am proclaiming

13:52 you, AI guru of the world, so I've, I've just stood upon you, this prominent position. You get to

14:02 make all the rules that we have to live by, flush it out. What, what are the rules? Give me some basic rules you're going to put in place. So you're going to age limits on. on AI applications,

14:16 you get to set all

14:19 the rules. You are the drafter, as well as only signer on the AI constitution of the world. So one thing, right, I mean, this could be controversial, but one thing I'd say is,

14:34 you know, instead of making a lot of, you know, specific rules, right, because, you know, it's very hard to make so many different rules But one thing I would say is, I would say if someone is

14:48 putting a product out there in the market, right, if someone is putting a product out there in the market, and if it has a negative repercussion on, you know, a certain person or a certain people,

15:00 right,

15:02 if you make them liable for that, right? And that's civil liability to it, yeah. Yeah, yeah, and then, you know, I think, I think then everyone who's. who's utilizing or putting AI out there,

15:16 right, or AI-based applications, they're gonna be thinking about, is it actually causing any harm, right? What I find, I think in some of the cases,

15:29 if you just put an information out there, piece of information out there, and it's coming from somebody else, and you're just using your platform to put that out there, it's okay You're just

15:43 providing a platform to put the information out there, and it's not your opinion. However, if you take that piece of information and you're magnifying it, or you're essentially, you know,

15:58 exponentially increasing the visibility of that by picking whom to show and whom not to show and how much to show, then you have, if you have essentially a responsibility in terms of how you're

16:11 doing it, that responsibility should not be taken. away as an example, right? So what I would say is that that responsibility and that liability, if you put it on people who are putting the AI,

16:23 you know, I think the world will figure out how to do it in the, in maybe, you know, more, it will never be 100 positive, right? But then I would think that it would be more positive than

16:34 negative. And I think, you know, that's the way to go. It's one. So I do another, in addition to Chuck Dates, he's a Java do a BDE, we call it the weekly summary of the energy business for

16:45 people that think Jim Kramer sucks. But one of the things we've done this year is, you know, this is the year of democracy. I mean, I think over a billion people are going to go vote. India is

16:58 having elections right now as we've speak. Obviously, United States election. I want to say it's 75

17:10 of the largest democracies are having their elections the share. And one of the things we've noticed, 'cause we're breaking down each one of these individual elections is it's scary, the impact of

17:24 AI. I mean, you're having video generated speeches by candidates that are put out there. They're not true. They're not true. Yeah. And it happened in the United States. It was Robo calling in,

17:39 I forget which primary it was one of the state primaries There was Joe Biden's voice. Yeah. Yeah. Yeah, exactly. Yeah. And part of me says, well, okay, that's really bad. We're gonna

17:51 regulate it, but the libertarian inside me goes, who are we gonna get to make that decision? You know? Yeah, I mean, that's always tough. And that's the toughest part, right? Like who decides?

18:03 And I'd say, I mean, you know, eventually they would get decided by the Supreme Court, right? 'Cause, you know, somebody is gonna, make a case and it's going to go to Supreme Court and it's

18:14 going to be decided one way or the other and it will set a precedent. But we have a system or every country has a system and I think,

18:26 essentially,

18:29 either it's going to be the court or it's going to be the elected body that's going to make those laws Yeah, I'm in

18:38 a breakfast club and one of the members of the Texas Supreme Court came and spoke to us this morning and it was fascinating because in the state of Texas, we actually elect our justices and that's a

18:51 little different than the federal system, you know, appointed for life approved by the Senate and it's rooted in the Texas Constitution, which says all sovereign power of the state is invested in

19:06 the people and the people in effect are allowing elected officials to act. on their behalf and we tried to put tight chains

19:17 on them because we were definitely afraid as a people, because this is post-civil war that those damn Yankees are going to come down and try to, because we weren't as southerners, we weren't

19:29 allowed to hold office, but we got to write our constitution. Yeah, yeah, yeah. But I mean, it makes sense. Ultimately, whether it's the justice in the Supreme Court or the Senate, there are

19:43 representation of what the people want in the country. And I think that's what defines, and that's what great about democracy, because if the society wants it to be a certain way, then it goes in

19:59 that direction regardless And I think there's no absolute right or wrong. There

20:07 always is dependent on the context and dependent on the society and the culture. and they're always straight offs. Exactly. Yeah, exactly. There's no perfect answer. Yeah, it's gonna be a

20:16 fascinating watch, watching how this plays out. So let's

20:22 go back to industrial world. Yeah. So what are you doing with your life that makes you

20:31 go, man, I wanna go be a software entrepreneur.

20:36 I'd say, you know, I always enjoy building things, right? My dad, he was an engineer, he's an engineer, and he basically managed some of the water treatment plans and purification plans, et

20:51 cetera. So I grew up watching some of the heavy industries. And where did you grow up? I grew up in India. Okay. Yeah, and so back

21:03 when I was growing up, I always enjoy that engineering side, just tinkering and breaking things and putting things. I think there is always something special about building something from scratch,

21:17 right? Or making something and something that can help others and makes an impact.

21:24 And I think that was always something that stayed. I mean, I did projects where I worked on cars. I basically did projects at home. And I think that's also something that's great about US is

21:35 essentially there is that culture of building things and doing things from scratch. So, and that's what I would say, always drove me to start a company, right? And then,

21:52 I was always looking at ideas or what is a problem that I'm passionate about that I want to really dig in. 'Cause it's a journey that you have to be passionate about. And before uptime AI, I was

22:07 working at, spend about 20 years more than 20 years. working in the industrial analytics or industrial application space, specifically about applying analytics to industrial problems.

22:19 First 10 years doing pure analytics for all of it from automotive aerospace to far more than oil and gas companies. Next about eight years, I worked at GE, managing some of their product lines and

22:36 focusing on AI and deep learning based solutions for oil and gas So I saw how challenging it is to manage operations in a plant. It's a very complex environment. There is safety aspects,

22:50 significant safety aspects involved. It's a harsh environment. And it's also very hard for people to just understand, people who haven't been to those kind of a refinery or a chemical plant to

23:05 understand what it is like there

23:08 And I worked a lot with. I'm senior experts as well, who have, who have lots and lots of stories. I mean, we have a chief subject matter expert in our company, right, who has over 40 years of

23:20 experience, you know, in the industry advised over 150 different plants across the globe. And so I've heard all these stories, right? And, and about the difficulties, the challenges, how they

23:35 solve the problems And that, you know, problem in my interest in building something really connected together. And so right after GE, that's when I decided that, okay, this is a problem that's

23:48 worth solving and making an impact on the entire manufacturing industry. And that's how, you know, I jumped into it. Because what you just described is way more articulate way

24:02 of what we try to do around here at Digital Wildcatters is we're a little more folksy about it. It's like the great crew changes happening. Yeah. And all this vast information about even something

24:14 as simple as drilling a conventional vertical hole. Yeah. When that guy dies, it's gone, right? Exactly. You know, being able to knowledge, capture, and then transfer to other folks is so

24:30 important. And it just feels like in the energy business we're 15 years behind the rest of the world. Oh yeah, absolutely. In terms of doing it. Absolutely It's funny you mentioned that the crew,

24:41 the old crew, I mean, there are different terms, like the gray hair and silver tsunami, right?

24:49 And the thing that really stuck in my head, you know, this is a, this is a, probably this is my, what motivated, what was the crucial point for starting up time? I would say is, you know, I

25:03 was actually in the airport flying back from London to the States. And we just finished a workshop with one of the oil and gas companies in London, and I was flying back with one of our SMEs, right?

25:18 And we're sitting in the airport and our SME gets a call and he tells me, Jack, I can't fly to back to the States. I gotta fly to Malaysia. The client is having a problem in one of the plants and

25:33 they're asking me to fly in immediately. And I know the client in Malaysia and I looked at the SME and it called him, Look, I know they have all the digital tools, right? In the market, they

25:48 have best-in-the-breed solutions, digital tools. Sensor on everything. Sensors, et cetera, right? Why do they need you, right? And why so originally? And he looked at me and his myelin said,

26:01 Jack, your tools tell him that there is a problem. I'm the one going there to tell them what's causing it, how to fix it. And you know, even better, I'm gonna do it even faster the next time

26:16 another client calls me. So I'm much smarter than anything that you have. And that was a lightning bulb moment, right? Or light bulb moment, right? Right. And you know, that's kind of stuck

26:32 with me. And after LFGE, you know, or Baker Who's GE, then I was thinking about different ideas, and this was one thing that really stuck in terms of how the

26:46 industry, right? It has a lot of thermometers and blood pressure machines. They kind of give you indications of where problems are, which is great. I mean, there's value there, right? But in

26:58 order to take and leave frog to the next level of efficiency, right, we have to put that expert.

27:09 into a computer and make it available for every young engineer who's trying to run these plants. And that was essentially the goal for starting up-time AI. Well, and you know, the other thing you

27:22 have to do, and part of the reason I'm a libertarian is at a very, very young age, I was fascinated by the assassination of John F. Kennedy in Dallas. I've been to Dallas a million times. If you

27:35 and I ever get to Dallas, we're walking around the grassing old for three hours until you finally say, I'd give Chuck. I don't want to have to. So I've always been incredibly cynical and I'm gonna

27:46 say this, part of what you just described, the other thing that needs to happen is that expert needs to be challenged too 'cause so many of the biases of 30 years of doing something a certain way

28:03 was based on data, and you were only able to measure this much of the data, not this much, and so I think the

28:14 next step of that is, okay, we gotta really make sure that expert is right. One of the things I've noticed is a lot of times when we'll take over operations of a field and you literally change the

28:30 incentive structure for a pumper from, I want so many barrels of oil a day to giving them their own little profitability statement. They always run the pumps slower. They make a little less oil,

28:43 but they spend way less on upkeep. Yeah, yeah, yeah, yeah, yeah. And everything, yeah, so. That's a great point, you know. I'll give you an example around that. You know, there was, you

28:54 see today what we have in operations, right, whether it's refining or chemical or whatever it might be. There are, you have different functions. You know, for example, you have reliability team

29:08 typically consisting of mechanical engineers. You have electrical engineers who are monitoring electrical equipment. You have chemical engineers doing process work, right? And so you have all

29:18 these different functions, right? Now, the profitability of your operation is not going to be based on just, you know, keeping it, keeping the machines running all the time. Because, you know,

29:29 you can keep the machines running all the time. But if it's consuming more energy, more raw material, more fuel, it's still not proper. Sure, for an equipment time, yeah. Exactly. So the

29:41 point is, if you want to really improve the profitability, you have to look at it holistically, right? Which is exactly what you said. You just can't run it all the time to pump, you know, more

29:52 oil. But, you know, you have to actually look at optimizing it, right? And so the challenge, though, is when you haven't, we have a case where a customer came to us and said, you know, that

30:03 that team was reliable to team. It told me that, hey, we have a pump that's failing every nine months, you know, and we want to resolve that, but you know, we replace it and it still keeps

30:16 failing and there's nothing wrong with a pump. So we said, okay, you know, let's look at it as a system and let's look at what's around it and then we took all the data and the sensor information

30:28 around, you know, not just a pump, but the reboiler, the tower, because it was a bottoms pump in a tower And we put it in the application and we showed them that, hey, it was essentially a

30:40 reboiler that was following, which is causing the causing polymerization in the tower, which is making the fluid really viscous and thick. And that was essentially what is causing the pump to fail.

30:55 And then the mechanical engineer was super happy. He said, you know, now I know it's not my problem, it's actually a process engineer's problem

31:05 But the point is, you know, it's unless you're trying to look at it holistically, right? You're not gonna go to the next level of efficiency 'cause we have reached a stagnation point where every

31:19 function is using their own little tools, right? There are point tools for different functions, you know? And they're using those and it's great, you know, for whatever, what they have been

31:30 doing so far. But if you wanna break that ceiling and you wanna achieve at next level of efficiency and operational excellence, you gotta look at them, you know, together. And typically the

31:42 reason why experts are counted in gold, or counted as gold is because they are typically the ones who do that kind of work, right? And that's what, you know, we want the application to do by

31:53 connecting those dots. Yeah, I'll tell you my favorite kind of data story because,

32:00 you know, you've been talking expert. And I totally get, we need to capture and all that. But the other thing too, with the level of detail now and data that we have and just the computing power

32:11 being able to look at it, one of my favorite stories that I'm going to kind of change the details to protect the innocent, but one of the large oil and gas companies hired a data scientist from like

32:23 HEB, you know, someone who knew nothing about oil and gas. And I think this person was maybe 25 I think I know where this is going, but yeah, yeah, well, and so anyway, one of the things they

32:35 wanted to look at was expenses, basically surrounding the pumpers truck costs, all this sort of stuff. And she studied the data and came back and just said, you know, if a pumper will never put

32:49 their truck in reverse, it'll, I forget what the stats were, it was like, it'll drop operating, he spends a 75 and it'll drop rex 44 or vice versa, whatever it is. And, you know, the old

33:02 pumpers like, what the fuck are you talking about? Hell, that ain't right. So anyway, they just redesigned, you know, a couple of pumpers routes. And you never went into reverse and the

33:13 numbers were almost exactly right. Right. Right. Right. It's amazing how just looking at the whole with just pure data. Exactly. I mean, you know, there are, like you said, there are ways

33:25 that are used to working. And there's a reason for why, you know, people work the way they do today, right? But it doesn't necessarily mean that there are better ways, you know, the constraints

33:37 that people had when they designed the systems, the way they're using it now. They might not be there today. And I think, you know, many of these technology tools are helping them figure out what

33:50 those, you know, the new normal for operation could be. Yeah. Yeah. Yeah. So you start this company Yeah, give me some. Crenereal battle stories about starting a company Crazy craziest thing

34:05 that happened that you never would have thought when you started a company I don't know if I can say

34:15 We started and then you know covid hit Come back count as you

34:23 know So we started end of 2019 right and Within the company we have Our technology team is located in Bangalore and we have commercial teams in you know in the ue. Today we have commercial teams in

34:39 the us. You know Canada Middle East etc. But Back in the day when we started it was just you know Two people me and my co-founder my co-founder lives in in Bangalore and we know each other for close

34:52 to 30 years believe it or not And so I I went to India and I was in Bangalore for the first time. you know, a couple of months, you know, trying to set up a team and starting to essentially, you

35:06 know, do some RD and build the application, et cetera. And then I came to the US. I think in February or 2020, after a couple of months. You brought it, your patience zero.

35:20 And then I was actually at a conference in San Diego. And

35:27 then, you know, after half a day, right? The organizers came in and announced it's done, it shut down, go back, right? And that was two months or maybe three months into the company, right?

35:42 I mean, you know, it was essentially, we didn't know what was going to happen. Because nobody was taking, you know, new initiatives. It was two weeks, right? I mean, that's what we're

35:52 talking about at first, yeah, exactly, flatten the curve two weeks. Exactly. And then, you know, it slowly started increasing. But, you know, It's

36:03 always one thing, I always look at it as, there's always ways to work around things and kind of figure out what's the right thing to do at that time given the constraints, right? And so we looked

36:16 at things and we said, okay, I mean, you know, this is the time a lot of people were laid off, right? And we hired good talent,

36:25 everybody was working remotely and we put our heads together and we work, you know, constantly, you know, try different techniques, different approaches, you know, and we spent almost two years,

36:38 two and a half years in the garages to come up with a solution that would really work, but not only working, but also, you know, the speed of deployment. We were not, we didn't want to build a

36:51 data science platform, right? We were not building something where we were gonna give it to a software engineer and say, here's all your platform. can build whatever you want to build. You know,

37:02 we wanted to build it for the people who are actually in the plants, in the trenches working, right? We wanted to give them a solution and that essentially meant that, you know, we wanted to give

37:14 them a solution that would be ready in three weeks for a full unit because otherwise, you know, today, what you'll find is you get, you know, if you want to do something more advanced, the

37:29 company will have to get, you know, maybe software engineers and say, let's build it ourselves. And it can take up to, you know, whatever six months, one year or two years, we have, you know,

37:38 clients who spent like three years before trying so many different things. And then, you know, eventually figure that it would be so much easier without timing, right? Because, you know, it

37:47 takes three weeks to deploy Yeah, yeah. Yeah, it's a love here in that story. One of the things I think we often overlook in life is

37:59 We talk about how when you're giving a

38:02 fresh slate or clean slate on something, that's when you're most creative. And I don't believe that at all. I actually think when you're boxed into the corner, that's when the creativity come out.

38:13 Absolutely. You know, Samuel, I think it was Samuel Beckett that wrote the playWaiting for Godot. He actually learned French to write the story because the constraint of learning a new language

38:29 and not kind of just resting on colloquialisms. Right. I think made him write a story that said exactly what he wanted to say. Right. And it's totally true. I mean, I would say sometimes I also

38:45 feel that. I think, you know, even I started, right?

38:50 Too much money is also not good, you know, because the constraints actually force you to come up with innovative approaches in solving problems and that exactly is what is gonna drive that push to

39:05 make it better and better and better, right? So yeah, I completely agree with that. So what, give me a good initial story of like your first sale or your first implementation or something like

39:20 that. I mean, how does that discussion go? You know, I'm Mr. XYZ Plant Manager

39:27 No, that's, it was essentially a great experience. You know, I, even though my experience, I did consulting, right? I did consulting for about 10 years and then I managed products at GE for

39:41 seven, eight years, but I was not necessarily, I was not into pure sales, right?

39:47 I was never into pure sales. So as a founder, you're going and you're essentially talking to a prospect or, you know, And I've always, one thing that was very clear from the very beginning is, I

40:01 mean, sale is a way of, you know, you solve problem for someone and you get a portion of the value that you're providing, right? And I always wanted the company to stay true to that, right? And

40:17 so we were very upfront. We were very straightforward with our clients. We said, you know, the first sale that we made was we went to the, we went to the client and we said, this is our version.

40:31 This is what we want to actually, you know, we're developing, we know how we want to do it, but this is the idea, right? We told them we don't have any other clients, right? This is the first

40:45 one. And, but the idea was compelling for that company, right? for we presented it to head of, you know, operations for one of the one of the companies, power generation companies. And they

41:00 said, you know, you know what, if you do this x, y, z, I'll see that, you know, you're good for it. And then, you know, we'll essentially go for it. So we said, okay, you know, it's a

41:13 challenge. We'll take it. And, and we essentially took that challenge. We did it. And we did it extremely well. The problem that they gave was was there was a major failure, right? There was a

41:25 major failure in the plant. And after that failure, they went to a, the OEM, right off that equipment. They said, you know, this, this major failure costed us, you know, millions and

41:37 millions of dollars of loss. And they asked, you know, OEM, what just happened? Because we have so many other equipment and we don't want any of those fail like this, right? And the OEM said,

41:49 there's nothing wrong with the equipment It could be in the grid, it could be something else, right? And then the company then went to an engineering research company and said, you know, what's

42:00 wrong with the grid? You know, EM is saying something wrong with the substation or wherever the transmission is happening, right? And then they looked at all of that and they said, you know,

42:10 nothing wrong here either, right? And so the head of operations, what is asked us is if you can figure out why that's happened and when it happened, I'm not gonna tell you where and what. But if

42:27 your application can figure it out, then I'm gonna work with you, right? And we took that challenge, we found it and the application was able to find it, you know, like four months before it

42:41 actually failed. And after that, they just deployed it for their whole unit. That was the first sale. Oh, nice, I love that. Yeah, and it's interesting that we also ask our customers, even

42:58 after the signup with us, we ask them, Why did you sign up with us? This is a space where a lot of big companies are there and a lot of solutions are in the market. But we ask them, Why did you

43:11 sign up with us? And one of the most common things that our customers always say is trust.

43:22 They say that you didn't tell us that you're going to do everything. You said you're going to do these four or five things. And whatever you said you're going to do, you did it extremely well and

43:37 you did it in a way that nobody else was doing. And I think, and I always say that we want to be respectful and make sure that there is value that is being provided, we're solving the problem, and

43:50 we're truthful to what we're doing. And I think, you know, the revenue and the brand and the sales are gonna happen. Yeah, no, that's cool and that's really good advice. So how'd you finance

44:04 the business? Did you go out and raise VC, friends,

44:09 family? How'd you do that? Give us a little color there 'cause everybody listening to a podcast with an entrepreneur always wants to go, how'd you get the money? Yeah, yeah, yeah So, when we

44:21 initially raised friends and family, right? That's how we set up the shop and we kind of, you know, started essentially hiring people. I talked to a CEO, an entrepreneur yesterday, I added

44:32 another F, friends, family and fools.

44:38 Yeah, yeah, yeah. But, you know, we have a, my co-founder, he actually, this is his second startup, right? So, his first audit was pretty successful. And.

44:51 There were friends and family who invested in that and, you know, so he showed how, you know, it could be, it could be positive. It could be beneficial. And of course, you know, we have

44:60 reputation within our circles that, that people were willing to trust us, right? So first was friends and family, which was a small round, which was a safe round. And after that. What's a safe?

45:15 So safe is like, you don't actually put a valuation on the company, because when you are so young and you're collecting or you're essentially, you know, getting the money from friends and family

45:27 who are not necessarily institutional investors, right? They don't know how to value the company. They're basically trusting the founders. So what we do is we make it as a safe round. What that

45:38 means is

45:40 there is a, there's a bottom line it like a catch. There's a net where if if the valuation goes down, it can't really go down below a certain limit, but there is upside. If someone comes in

45:54 tomorrow and puts a value on the company and the valuation is even higher, they still get at the net. Essentially, what it means is there is a protection for the bottom side, on

46:08 the downside, but there is benefit on the upside. It makes it simpler When you go to a VC, when you get the money and they put a valuation, the safe actually shares would convert at that valuation.

46:25 The reason I asked is we've raised some money. I think we've raised45 million today at Digital Wildcatters. I want

46:36 to say the first 2 million of that was a safe. It's basically, hey, if the company blows up, you get the money back first. It's kind of like a note in that, but at the end of the day, we'll

46:49 discuss valuation when we have a pricing round. Exactly. It's exactly that. So you did that, but then you raised institutional capital later? Yep, we did. And we raised it in the middle of

47:01 COVID, because we didn't know how long it was going to happen. So we raised it in the middle of COVID about May, June of 2020 That was the first, I would say, pre-seed, right? And then after

47:19 that, we raised down the round last year, or yeah, end of 2022.

47:28 And that was our seed round, right? And now we're basically in the process of closing for a series A. Nice. So give me as many as you want, but at at least one or two bits of advice for a. a CEO

47:44 who's out trying to do what you just did. Yeah, so one of the things that is important to understand is fundraising is a, it takes time, right, dedicated time. And the best way to do it is to be

48:05 prepared, right? And by preparation, I mean, you're trying to convince someone that this is a business that's gonna grow to a hundred million over, you know, five years, six years, whatever it

48:15 is, right? And the other person doesn't really understand exactly what you're doing, right? I mean, they may have general idea. So they were gonna ask for notifications, material, information,

48:30 right? And you as a founder have to provide that, you know, that kind of information and knowledge and convince them So that requires some collateral, some material, right? whether it's about

48:43 competition, whether it's about your own company, whether it's about revenues, whatever it might be, right? So the first thing I would suggest is, make sure that you have all that collateral

48:54 prepared. Number one, second, not every VC is gonna invest in every type of business. Second, like you're on the energy space, find institutional investors

49:06 who have a focus in the energy space, find investors who focus on the specific market that you're targeting, right? And then you create a list of all the VCs, who you think are good for the

49:19 business that you're building. Third, I would say, then make sure either you have contacts or how can you reach out to the VCs, right? So that you have a warm connection instead of throwing in

49:33 just a random email. I'm not saying don't throw a random email, but you have a better chance of getting your time do we see when you have a warm connection? So those, I would say, are if you do

49:46 that well, then you're basically gonna be very efficient in that process. 'Cause one of the challenges for a founder when they're raising the money is you lose a lot of time focusing on raising the

49:59 money instead of building the business. And that's huge. I mean, and so - I think that's been Colin's big realization about fundraising. He's just, holy crap, it takes a lot of time It's a

50:03 full-time job. Exactly, exactly. One other thing

50:06 I'll say

50:08 about fundraising that

50:19 I try to tell entrepreneurs, and you tell me whether this is good advice or crappy advice. But one of the things I see too often is a founder is trying to, and I hate to be, pardon my French here,

50:34 but they try to get laid on the first date. And I keep saying no, this is like dating. You want the next date. you can sit there and have a fight about some arcane detail and try to prove the VC

50:49 or you're correct. That's great. You may have won the battle, but you lost the war 'cause that person's gonna be next, you know? It's, you're trying to educate enough so they'll invest more time.

51:01 'Cause ultimately at the end of the day, they're gonna have to spend a lot of time with you to cut it, check. You just want them to spend more time with you. Right, right. I completely agree I

51:11 think it's like a sale, right? I mean, you

51:17 can't just convince someone about something that you've built over years and years of knowledge or your understanding of the market. And you can import all of that knowledge in a single call. It's

51:29 very difficult, right? And you have to be patient and you have to explain why this is the best thing that you're doing and how this is gonna create an impact, both from a revenue perspective as

51:39 well as from

51:41 a society perspective. And I think it's important. I completely agree with what you said. Yeah. So tell me who of all the vast majority of the world that listens to this podcast, ha, ha, ha.

51:57 But of folks listening to the podcast, who do you want picking up the phone and calling you or emailing you? Who's kind of, if your customer is everyone in the world, how do you narrow that down?

52:12 Yeah, yeah. What I would say is any manufacturing operations, right? People in your companies who have manufacturing operations, especially in the heavy asset space and more specifically in the

52:25 process industries, right? So for those folks, if you're looking to improve your operational excellence, right? Either you've stagnated, either you are trying various types

52:39 solutions, they are not giving you the results. It's taking too long for you. Your operations teams are not accepting. You can't find the ROI because, you know, your management doesn't see the

52:51 return after two years. Those are the folks, you know, call us. You know, we will show how we can deploy in, you know, in a matter of weeks. In every customer where we have deployed, we have

53:07 a record of 100 success with scaling up within almost within a year. And so we will show how we do it. And at the end of the day, I mean, you know, we want to also help the industry, right,

53:19 kind of see what is available in the market. And we always tell them, you know, at the end of the day, if it fits their needs and if it fits their, you know, business objectives, we're happy to

53:31 work together and see So what's kind of a ballpark that the software may cost? you told me100 million, that would surprise me. If you told me75,

53:45 000, that would surprise me as cheap, but ballpark, what do these solutions look like? Yeah, so what I would say is it certainly will depend on how big the plant is, what are we covering, right?

53:59 But typically, most of our clients, you know, anywhere between, you know, let's say 150, 000 to 300, 000 is essentially where they would typically start, right? And go from there, right?

54:12 But it always depends on, you know, the size of the operation. Yeah, gotcha. So how do people reach you? Website, you LinkedIn, Twitter, how do people run you down? Yeah, I mean - Email,

54:27 whatever you wanna give. Yeah, yeah, so we have, our website is uptimeaicom,

54:33 right? It's not uptimeai, It's actually UptimeAIcom.

54:38 because people are always confused with that and there are essentially, you know, we have, there's an option to request for demo, you know, find people there, you know, there's a chat, you can

54:50 always, you know, chat with someone on the, on the website as well, but you can also reach out to us at info, info,

54:59 infouptimeaicom and certainly one of our team members is going to certainly reach out to them. Cool. Well, JAG, I appreciate you coming on. It was nice surprise to just have something dumped on

55:11 my calendar like this. No, it was, it was a lot of fun. You know, I really enjoyed the, the range of the conversation was really, really good, you know, all the way from AI. I didn't take my

55:21 ADD medicine this morning, so what the hell? Yeah, it was good, it was good, yeah. Thanks for having me on the show. to be a pressure. Absolutely