#378: Sergey Rekutin on Building a Business Through the Ups and Downs
In this week's episode of AU, guest Sergey Rekutin of Cravtsman joins Ron to discuss how Cravtsman nearly closed in 2023, and the pipeline and expense discipline that turned it around.
This week's episode of Automation Unplugged features Sergey Rekutin, President of Cravtsman, a high-end residential integration company based in Englewood, New Jersey, with projects across Manhattan and the Hamptons.
Sergey started Cravtsman in 2008 as a side business that grew out of his job at a cable company. He is an active member of the Azione and Cinergy communities, and he has spent the last several years building a reputation for design-driven work in some of the most demanding homes in New York. Today he is sharing the lessons from the ups and downs along the way.
In this episode, Sergey and I discussed:
- How Cravtsman nearly closed in 2023, and the pipeline and expense discipline that turned it around
- How he uses AI to build dashboards and custom tools that connect his software, while leaving the decisions to people
- And His vision for a "headless" future where a personal agent travels with you and every home runs its own MCP server
Whether you're growing your integration business or working to steady it, Sergey's honesty and advice will give you something to take back to your own team. So settle in and enjoy my conversation with Sergey. Let's get started!
SEE ALSO: #377: Juan Camou on Why AI Starts With Your Data, Not Your Agents
Transcript
Ron:
Hello there. Ron Callis with another episode of Automation Unplugged. I hope all of you are doing a little bit better than me. I'm fighting back a cold, but I'm on the better side of it. Fortunately for you all, as always, we're pre-recording the show, so if I need to step aside and sneeze, you will never know it, because we will edit it out and you won't have to see it or hear it. But my poor guest will, so he's already earned some credits there. He's a kind and gentle soul, and I know it's not going to bother him too much. Or at least that's what he tells me. So let's go ahead and jump right into it. Today we have a great operator, and like many of us who have been at entrepreneurship and business for a while, he's had his ups and his downs, and everything in between. He is game to share some of those lessons with us. Many of you out there in Azione will know this person. In the Cinergy group, you'll know this person. He's an active member of our community in the custom integration space. So today we have Sergey Rekutin. He is the president of Cravtsman, and they're based out of Englewood, New Jersey. I'm excited. The joke is I've been asking him to be on this podcast for many years, so I see him chuckling down on the screen. He and I ran into each other at CEDIA and he said, "Ron, I'm ready." And I was like, "I'm ready too." So here we are, recording a show together after many years of business and relationship. I'm excited for you all to learn more about Sergey and his business, and let's get this party started. Sergey, how are you, sir?
Sergey:
Hey Ron. How are you doing?
Ron:
I am good. I'm here in Florida, and the nice weather has finally broken out in South Florida just in time for me to get a cold and have to stay indoors and try to recover. So that's how I'm doing. How are you doing?
Sergey:
Since we're talking about weather, it's been pretty good. It's been raining for three or four days, and now we finally have a break. It's nice and warm, which is very good.
Ron:
That's nice. Where are you coming to us from? Are you there at your office or your home?
Sergey:
I'm in the office in Englewood, New Jersey. It's a couple of minutes away from New York City, just across the George Washington Bridge.
Ron:
Okay. Give us a little bit of a download about Cravtsman. What type of business is it? What type of projects do you do? Where do you do those projects?
Sergey:
We've been around for 18 years, about to be 19 very soon. We had a long-winded path. For the last eight years or so, we've been focusing on high-end residential, and it's been climbing into more and more high-end space. I like to call the houses we work in ultra high-end. That's typically where we operate. We normally work in New York City and the Hamptons territory, and the other boroughs of New York City, but most of our work is in Manhattan itself and some of the Southampton territories.
Ron:
Okay. What's in a typical project? What type of technologies? If you want to mention brands, you can. You also can leave those out. But what's a typical project like?
Sergey:
Our specialty, and the reason why these architects and interior designers work with us, is because we're extremely focused on the aesthetics of the technology. So we sell a lot of Meljac, and obviously Lutron behind it. The clients we work for would spend a lot of money, and when I say a lot of money, it's multiples of hundreds of thousands on things like door hinges and doorknobs. They obviously don't want to see anything plasticky or artificial-looking on the wall. That's why we go to natural materials. We do all sorts of control, but our go-to way to approach projects is how do we make this the most eye-pleasing way to implement tech. This is why we get hired, because we do a lot of motorization. We hide things behind moving things, moving walls and panels and-
Ron:
Hmm.
Sergey:
things. We research constantly new materials from which keypads could be made and bring different companies into the market, sometimes not very popular. We do a lot of extremely custom stuff, meaning we would take a standard solution and modify it to fit the specific need.
Ron:
You have to give me an example. An average company may just offer X keypad or whatnot. You guys go the extra mile. What's an example of going that extra mile?
Sergey:
The extra mile goes like this. This is my favorite example. I absolutely do not recommend anyone do this. This is not a good business.
Ron:
Yeah, yeah.
Sergey:
But we had a project, and it took five years to build, and it's my favorite project. There were mosquito nets on the really large windows that have a skirt where the track is sitting. Through years of conversation, somebody forgot to mention that the track needed to be a little bit taller. So they missed two inches of the track, which is an extrusion that they have to make. When we installed the shades behind the track, the client was actually able to see the white line when the shade rolled up. It's a black fabric, a dark fabric, but when it rolls up, it's white on the other side, so she saw the white color. Again, my recommendation is not to do this, but we found rubber paint, and we painted the back of the shade. Then she realized that once the shade is down, she also sees a little bit of a white strip anyway, so we had to paint the top of the shade a second time. So when it's all the way up and all the way down, it actually blends in completely. That's the extent that we would go to. On the same project they bought a bunch of lighting fixtures from Europe. We had to retrofit all of them in order to properly work in the environment.
Ron:
Wow. From an electrical standpoint, you had to rewire them?
Sergey:
Yeah, we had to slightly rewire them, change power supplies, and so on and so forth. That's the extent that we go. And when it comes to things like that, we're not afraid to talk to Meljac and have them come up with a new finish that would match what we're trying to get to. We'll spend the time, and we know how to do this quickly enough not to delay the project with the small nuances. We have a sample coming in soon. We found this person who does glass enamel artwork, so we're trying to make some copper enamel keypads for very special clients. The cool thing about it is that it never fades. It's one of those things that has color and never fades. Those sort of things we do all the time and we like to do. I feel like our general capacity to do very custom work is much higher than a lot of other companies, where for them it's a special project, but we do so much special stuff. For us it's very normal.
Ron:
It's your normal state.
Sergey:
Yeah.
Ron:
No, that's great. You and I ran into each other, and it actually led to us being able to spend a couple of minutes and coordinate getting you here on the show. We ran into each other at CEDIA out in Denver, back in early September. What stood out to you from CEDIA '26? What were your takeaways, or maybe the biggest value for you of being at the event?
Sergey:
Obviously everybody who's been there knows the show was smaller, and it felt like fewer people attended. But I actually found it extremely valuable because you could spend sufficient time with the people that you don't see very often and catch up with them.
Ron:
Mm-hmm.
Sergey:
Make sure the relationship is still live and well. We ran into each other, and we saw Julia, who is a hard person to catch on this side of the pond. All the people I enjoy interacting with are typically busy because of the cohort of viewers passing by all the time, and they want to touch as many people as they can. This one was extremely beneficial. It felt really good. The schedule was not that crazy. I feel like I had so many fantastic conversations. It just felt very good.
Ron:
I'll echo that. In terms of expectations being met or exceeded, or you're let down, it goes into what you were expecting out of the event. We didn't know what to expect from One Firefly. For us, the event was pleasantly busy enough. We only need to have so many meaningful conversations for it to convert. And for me it's reminiscent of the Lightapalooza show from the last two or three years. It wasn't such a crowded event, but the right people were there, so you're able to just have really thoughtful, meaningful conversations. 30-minute conversations, 45-minute conversations with people that mean a lot to your business or to you personally. That was my experience at CEDIA. I thought it was great.
Sergey:
Yeah. Since Tech Summit is sort of fizzling out, and at least in the New York, New Jersey area it wasn't that great the last couple of years, we don't really see our reps. And some vendors don't show up as often. When you're busy, you kind of welcome it, but it's really nice to talk to the people you work with all the time.
Ron:
Yeah.
Sergey:
Because they make your business run better, and it's nice to stay in touch.
Ron:
I think that to look at a show like CEDIA or a buying group event, and I know you're in Cinergy, to attend those events, I think you're approaching it from a really smart way, mature way, maybe lived experience, but it's so much about the people, right? It's so much about what we do. The tech keeps changing, but these are the people that you're going to work with, your reps, your manufacturers, your peers, and there's a lot of value in spending quality time with people. Especially in this age of AI and all this technology changing at such an accelerated rate, it's good to just see people and have good conversations.
Sergey:
I completely agree. Also, everybody, doesn't matter whether they're in a role at this company or that company, they circle around between the companies all the time. Typically there's not enough escape velocity to leave the market completely.
Ron:
Right.
Sergey:
So they would just go to another brand that you support. It's just good to stay in touch. You never know how it's going to turn out in a year or two. Good people are good people, doesn't matter where they're placed.
Ron:
I agree. Well, let's jump back and hear a little bit about your story, your background. Now I get to reverse things on you. Over the years you've told me you've enjoyed hearing about the backstories of a lot of the people that have been on this podcast, and now we get to learn about your backstory. Where do you come from?
Sergey:
I can take it as far back as you want, but I can do a quick history.
Ron:
Yeah, please.
Sergey:
Everybody likes immigrant stories. I was born in Tajikistan, in the city of Dushanbe. It's a nice remote place.
Ron:
For listeners, we're going to spell all that out in the show notes, because I don't know how to spell any of that either.
Sergey:
Basically it's a tiny little Soviet republic right outside of Afghanistan. Everybody knows where Afghanistan is. Well, at least they know the word. Then the Soviet Union started to fall apart around '89. Family picked up and moved all over the planet, and we ended up in St. Petersburg. So my adolescence happened in St. Pete's. Cultural exposure and so on, a very nice place to get that part of education. Then things hit a sour note around that time as well. The whole country was going through a very rough period. My dad, although he's a scientist, was in business, and he managed to start something in the States, which allowed him to move the family. But we got our status because he has several patents to his name. One of them was for rocket navigation, so my dad is a rocket scientist.
Ron:
Oh, wow.
Sergey:
To the point that we had an FBI agent come in, within maybe two months of when we moved, to debrief my dad. Even at that time, a Motorola flip phone had more technology than what he was working on, but still, it's very unusual. Those are very rare visas for people to get.
Ron:
That's crazy.
Sergey:
Yeah. Since then I've lived in New York or all around New York. I got my high school and college education in the States, a degree in computer electronic systems engineering, so I'm basically doing what I studied for, which is kind of nice. When I graduated, the economy collapsed because of the dot-com bubble, so there was no work for me. I went to work for a cable company and I was moonlighting as an admin in imaging offices, the ones that do MRIs and X-rays and all that. The guys that own those places made a lot of money, so they bought a lot of TVs all the time and different electronics. Since I was the person who was fixing stuff in their offices, I got invited to install TVs and connect those surround systems and whatever. That snowballed very quickly. I think the boom with the TV started in '05 or '06 with the flat screen TVs when they got cheaper. In '06, I started to do this on the side from my side job, and very quickly that took over my side job. Then in two or three years, I can't remember, I quit my day job in order to do this. We did the stupidest work on the planet. We were just hanging TVs, a lot of TVs. The company was opened in 2008. New York City was where we worked. We've been based out of New Jersey all our existence, but all the work is in Manhattan. The reason is that in '08, business in New Jersey died out completely, and this is where we got all our clients. Then over the years-
Ron:
Was this before the... Because right in '08, I empathize with you. I started in November of '07. But that was right before the Great Recession was announced. Did you put your shingle up before the world started melting down, or right after?
Sergey:
I so did not care. I was just doing-
Ron:
You were starting no matter what.
Sergey:
It didn't matter. That was such an incredible supplemental income for me, incredible money, to the point that the reason the company got started is because the original name for the company was 5163 Company. That was my tech number at the cable company. I named the company the same as my tech number, so if I answered the phone and my boss called me, they would pronounce it as my tech number and I wouldn't get in trouble, because it was definitely in conflict with the whole employment and doing this on the side. People would write 5163 on the check, and I would write my name next to it. At some point the clerk at the bank said, "Listen, I can't take these checks anymore. You bring so many of them, you've got to go register a company." So I was basically forced in 2008 to do this, and since then it's just kept going. I very quickly found people to work with. It was a whole big mess for maybe six years or so. We did a lot of work. I had partners. Partners wanted to do volume work. I wanted to do special work, because I like complexity more than the business operation of things. And-
Ron:
You were discovering what type of business you were. And in the beginning you were saying yes to, it sounds like, whatever you needed to say yes to in order to get revenue.
Sergey:
Right, right. But then they didn't like complexity, so we kind of fell out of each other's orbits. They left. I think the last one left in 2012 or 2013, and then I was the sole owner after that, and I started to press toward the high-end market. In 2018, we rebranded as Cravtsman and focused strictly on the high-end market. We keep getting into more and more high-end spaces since then.
Ron:
Interesting. Okay. Was it in 2018 when you rebranded to Cravtsman that you started positioning yourself for more luxury work?
Sergey:
No, we did luxury work for maybe four years before that.
Ron:
You did?
Sergey:
Yes. We would work for CEOs and the marketing executives, the ones that run large marketing companies on Madison Avenue. They would take time from their day and spend 30 minutes with me explaining how terrible my name was. They did not like 5163 Company at all. That did not jibe with anybody. So I thought about it for two years, and then this name just came to me. We rebranded as Cravtsman. By that time, we already did all high-end work.
Ron:
Okay. So take us to '18 when you rebranded. You ran a couple of years, then you ran right into COVID. What was the COVID high? I'm leading the witness here a little bit, because I know what happened on the other side of COVID, but take us through that.
Sergey:
Yeah. We started to do really well in '18 and '19, and then COVID hit. It was game busters, right? Everything sold. You could throw any number out and they would say, "It's no problem. Just do this. Come over, do it." Business really jumped quite a lot and I started to hire people, assuming we were going to keep growing, and we did. We grew from one thing to another. I had a lot of office folks now to process all the communication that's required to do jobs really well. At the end of 2022, I started to lose sight of things. I didn't notice that when I sent out a bid, it would not get accepted pretty often, even though there was a close referral or whatever. I was thinking that something else would close, and I realized in early '23 that we were actually very quickly running out of money with so many people on the payroll. It was a very rough time. Through the first half of the year I was thinking that I'm probably going to close the shop and just work for somebody else in New York City. The person I wanted to work for actually sold the company, so I was like, "Ah, missed opportunity there." By the end of '23, we finally righted the ship. I found good places to work for almost all the employees that I had to let go. Somebody found their own jobs, and I found a couple of jobs for the techs that were here, just not to lose talent in the market. In '24 things leveled out, and then in '25 we did better, quite significantly.
Ron:
How's '26 been?
Sergey:
'26 has been scary good. We set a very unrealistic goal for '26. We said we're going to double what we did in '25, and if we're not going to butcher the last quarter, it's going to come true.
Ron:
That's amazing. Congratulations.
Sergey:
Time will tell.
Ron:
Yeah, we don't count our chickens until they hatch, or whatever the cliche term is, but I get it. It sounds like things are trending in the right direction.
Sergey:
It seems to me the same way. We did this without adding more than one person to the list of employees, which is great, because we learned how to outsource some of the back office work overseas, and that helped a lot to keep expenses low.
Ron:
Well, just to zero in on that topic. In '23, and it sounds like maybe in '22 it started to happen, your expenses got out of control. What did you learn? Because you were able to pull out of that. Some businesses never pull out of that, and they go out of business. You pulled out of it, maybe by the skin of your teeth, but you pulled out of it.
Sergey:
Yeah.
Ron:
And you said you steadied things, and then you got back to growing. Looking back at that, what could you have done differently that maybe would have protected you or helped reduce some of that pain?
Sergey:
The pragmatic answer to that is, you need to monitor your sales pipeline like a hawk. You have to make decisions, you have to create probabilities and keep a very good eye on it, because the moment that number drops, and ours was very low, you should expect to have no work. In our industry, it's very easy to get a good year because a job closed. One job can make a year. But in reality, you have to be extremely focused on the fact that if your pipeline dries out, you have to start making adjustments. You have to cut down on people and so on and so forth. During the pandemic, I focused a lot on operational efficiencies. We had things like utilization information. We tracked whether people actually created those billable hours, and the combination of sales pipeline and utilization is a really good sign of where things are going. You have to forecast your invoicing as well. That's a really good sign, and I was completely oblivious to that. The only reason I might sound more intelligent is because I went through this experience. I think it's the only way I actually learn anything, is if I have to get in trouble first, and then when I get out of it, now I have all the wealth of knowledge. So I really did not pay attention. Everyone I respect a lot in this industry, every single time you talk to them, they always tell you to keep an eye on expenses. That's the only thing you really have to focus on when you're running a business and making decisions, hiring, and so on. I completely ignored it, and that almost shut down the company.
Ron:
I'm curious, today, in the last quarter of 2026, if you are using AI to help in the monitoring of those things, your sales pipeline, which then influences your cashflow forecasting for the year ahead. Do you do this with human manual processes, or have you layered in AI into any of this?
Sergey:
Decision-making is obviously human. It has to be. Otherwise it becomes... anyway, it's just somebody who has a lot of experience and a feel for the situation who needs to make decisions. AI is not good at that.
Ron:
Yeah.
Sergey:
But what AI is good at is helping us aggregate the information required to make those forecasts. We use a lot of tools that parse out information from different software packages. There are no good reports in QuickBooks that tell you if you're upside down or not. You have to tease them out manually and put them in a spreadsheet, and put your information over it. What we use AI for is to automate a lot of this. I hate doing it. When we started to implement little bits and pieces, we used the AI to create software that would bridge the gap between different software packages and give us dashboards. That's what the AI was used for. I don't have AI advising me on the decision-making.
Ron:
Yeah, and I understand that. On this podcast, once a month I've been doing a new AI Insights episode. For example, the show that just dropped recently was with an integrator using AI, not to make their decisions, but to aggregate and synthesize the data to give dashboard views. As I'm talking to more and more business owners, we're doing it here at One Firefly, not to make decisions for us, but to aggregate and synthesize data so that I, as the CEO and founder, and my head of finance and my head of sales have information in front of us that's custom tailored. That probably would have been possible before AI, but it would have cost 10 to 100 times more to develop those dashboards.
Sergey:
And the time it would take, and then the upkeep of those things. None of it exists anymore. If you understand what you need, you basically prompt it into existence. That bridge of having the software engineer on staff to do very specific things for you is gone. So we basically use AI to write tools that help the business function. We are slowly rewriting a Cravtsman OS in the background. Hopefully that's going to come online maybe next year, so we don't have to rely on off-the-shelf software, because the shop is still small, in order to adopt our custom tool. That's where we're leaning. Essentially it's a tool to write software for us for the most part. And then obviously Grok Bot is amazing. Sorry to plug it in.
Ron:
No, I haven't done Grok. I've been tuned in and I've been hearing people really love Grok Bot, and now there's the new Muse from Meta, Meta Muse.
Sergey:
Yeah.
Ron:
And then OpenAI just launched theirs, Dots or something, their personal agents.
Sergey:
It's pretty good, but...
Ron:
Grok Bot.
Sergey:
Yeah. It got to the point that in a week of using it, we removed half of the functionality of my executive assistant that deals with the emails and basic filing and keeping people informed. Now she can do much better and more important things for the business. So very, very excited about that. It's malleable, so as it goes it makes mistakes, but everybody knows how that goes. It's a really easy way to do it, and it's very manageable.
Ron:
What would you say to any business owner who's leaning in that direction but hasn't done any AI development? And I'm speaking in regard to not using AI to write an email. I'm saying actually using it to create tools and systems, or system improvements, for your business.
Sergey:
Dashboards are the easiest approach. You just say, okay, I have, I don't know what stack people have, I have D-Tools-
Ron:
Or I have QuickBooks.
Sergey:
Yeah. So I have D-Tools, and it has some information. I have QuickBooks that has some information. How do I see it in one place? Then just ask that question, see what comes out, and you'll be amazed. After that, it literally will take you half an hour on Sunday morning while everybody's sleeping, and by Tuesday you will have integrated most of your software. You'll have created the abstraction layer completely by then. So the only valuable thing that's left in the software now is a person who understands the process.
Ron:
Mm.
Sergey:
As long as you understand that you are now the most important part of software writing, go ahead and do it.
Ron:
I could go so many directions with this, but I'm going to keep it within our channel and ask, when does this level of programming or configuration come to home automation? I'm going specifically into your customers and the work that you're doing to configure that customer's home. Heck, you could even go the next step and say maybe one day that homeowner can talk to their system and change things. But for right now, I don't want to go too crazy and say-
Sergey:
This is my speculation.
Ron:
All right. Let's get our Magic 8 Ball here, predicting.
Sergey:
Okay. I feel like we're going in a headless direction, meaning we install systems that have interoperability and communicate with each other that have no UI. When I say no UI, nobody has any UI anymore, not a hotel space, no one. You basically walk into the space with maybe a QR code, and you have an agent on your arm, in your phone, that has a dashboard of your life. It has your most important emails, your kids' schedules, et cetera. It has a layout of how you prefer to look at music. Maybe you sort the music a specific way. Maybe you gravitate to this or that. You have your list of shows you like to watch, all of those Netflix things you're following. So you have those pages for entertainment. You walk into any space, and I don't know how it's going to do it, you just ask your UI, "Can I watch a TV here?" Then the AI will connect to the headless control, and your interface would be the control for that space. You play the sound, the music, the way you want it. Your interface essentially transports to the space that you're in, so you don't have to think about anything. It's still on your phone. It still works exactly the same way in your house, in the hotel room, in your second house, in your friend's place. It's just one interface for all your life. There's nothing else. You just move around with it and all the systems are headless completely. Nobody writes a useless UI, because no one is good at doing UX work, and you just make your own thing and you travel with it.
Ron:
How many years out?
Sergey:
The first people to do this, next year maybe. Mass adoption, less than five years.
Ron:
All right. While we're playing the future game, what changes in the life of the integrator as a result of this?
Sergey:
Well, I think the adoption of technology becomes a lot more prevalent, so people will keep installing TVs, and they'll keep installing networks, and they'll keep installing some kind of controls. Some of it maybe will be integrated directly into the TV. I feel like that's probably the way TVs are going. But lighting control, if you want, could keep a nice pretty keypad on the wall. It still requires a lighting control, but the lighting control is completely headless. So you install the system, and the person just says, "I have a lighting control in my house. Connect it." Now you have all the rooms and everything else working. I feel like our work is just going to get busier because of mass adoption, because everybody will want to do it to some degree.
Ron:
And you still need a professional to design and install it, and there's a certain segment of society that's absolutely going to want the white glove treatment to get that designed, installed, and serviced.
Sergey:
Yes, that's right. And on a tangent, maybe technology will be so easy that the builders will actually just do it. But you still need somebody with a deep understanding to configure some of the interoperability, which I don't think will go away. Even with the rise of MCPs and things connecting to each other exceptionally fast, you still need somebody who can design the system properly and configure it in order for that headless system to exist. Maybe the next step of our evolution, because it's a little bit difficult for vendors to pivot in these directions since they have departments and things like that, would be that our role is to provide an MCP server for the house that then allows other systems to be integrated into it. It's a unified interface for your life that travels with you. You go from one house to another, and your friends come in, and they all just connect to your house MCP server, and you give them permissions and things like that. So I feel like that's where we're going to be, very shortly.
Ron:
I agree with everything you just said. I have no comment other than to say no one has said it that way on this podcast. I think we're going to be referring back in time to what Sergey painted as the picture of the future. Not going to hold you to it, so you don't have to feel any obligation or weight, but I think it's a really neat vision of how things could land. It makes a lot of sense. I've heard so much hype about the power of the personal agent that most people in society, including myself, and I'm feeling pretty far down the AI rabbit hole with development, have been there since at least '23. But I still haven't done Grok Bot and I haven't done Muse, because all of that is so new in the last 30 days. I just feel like I'm a month or two behind. Yet I listen to some AI podcasts and various content, and people that have gone down that slippery slope say, "Man, there's no going back. It's too fantastic."
Sergey:
Yeah, it's too fantastic. You can see all of that stuff coming from the depths of Hacker News. It percolates up, and next thing you know, big companies release something that is very user-friendly. Grok Bot is, Claude Bot, OpenClaude essentially, like a transplant, right? Or Harmonis or something. Muse is the same thing. It just very rapidly becomes-
Ron:
Productized.
Sergey:
Yeah, in a way.
Ron:
Commercialized.
Sergey:
A lot of adoption hurdles get removed because these big companies are jumping on it, because they see the value. It's just going to go at crazy speed in that direction.
Ron:
It's going to go viral. The numbers already on Muse, I think I heard this in the last week from one of my podcasts, and don't hold me to this, folks, but I want to say that it reached something like 340 million downloads in record time. The fastest of anything ever in human history, and they compared it to ChatGPT and all the other prior records. It was like they lit a fuse, and the fuse is going really, really fast. Apparently the product is really, really good. What's even weirder is if you go back and look at the news from as recent as June of '26, people were talking about how Meta was so far behind with AI, and their staff was quitting, and their leadership was changing, and they were falling apart. And now, three months later, they have the hottest product in the world. And you're like, "How is this possible?"
Sergey:
Because they own inference. They have so much capacity that they can do things like large scale, very simple tasking in a very easy way. The AI is maybe not the best at solving millennial problems, but they have the capacity. That's what Grok is doing. Grok is not the best model out there, but it's really good, and it's got the compute. They do have an abundance of compute, more than anybody else, and then they can do this thing like Grok Bot, which for me was elevating. Actually my days are now productive while I'm driving from appointment to appointment, because I just talk to it, and it does all the things, like emails people and so on. I don't have to stare at the phone, and I just command the whole thing. If I feel like it's a repetitive task, we figure out a way to do it, and it just does it. It's so good.
Ron:
It's a new world. All right. I'm going to ask you one more futurist question. I bring this up every so often to gauge people's feelings and opinions. Humanoid robots. Do integrators deliver them? Do they sell humanoid robots? Do they service humanoid robots? Because I think a humanoid robot is an interface in the home. Is it not going to connect to all the lighting and the shading and the audio video systems and the Wi-Fi, and simply be an interface into the home in the not too distant future? And then as an integrator, are you not in a very interesting position to be in that value chain?
Sergey:
Sorry, I thought about this literally nothing. I do not think about robotics programming.
Ron:
All right, we're getting you raw. Raw, unscripted, zero processing.
Sergey:
I mean, the natural evolution of AI is to interact with the physical world. It interacts with our computers. The next step would be robotics. That's naturally the way to do it. So would we have some kind of robotics in our lives? For sure, 100%.
Ron:
I plan to have it in my house washing my dishes within the next 12 to 24 months. I'm just saying. I promised my wife, "We're going to do this," and might even get it vacuuming as soon as it's ready.
Sergey:
Maybe. I don't know. For me...
Ron:
Does that weird you out, that idea? Are you on the "not on my watch" side?
Sergey:
No, it's just that I'm very terrible at predictions for physical manifestations of technology. I did not think for the longest time, and we're talking about 25 or 30 years ago, that we were going to have self-driving cars. Now we technically still don't have 100% self-driving cars, but it's extremely close, to the point that if I had bet money on it, I would already have lost it. So this is definitely happening. It's hard for me to prognosticate in that direction. But something will happen. It is such a different problem, because the roads have a structure to them, and even without the structure it's still somewhat flat. Robotics operate in 3D space, so it's a much harder problem, with people interaction and so on. But we'll see. It's not outside of the question. Whether the integrator will be the person to deliver robotics, I absolutely don't think that would be the case.
Ron:
You don't think the integrator will be in that equation?
Sergey:
It would be the same thing as Tesla. You go on your phone, put in the order. They make it in, I don't know, five or six months, it arrives at your door, says hello, and now you have the thing. Nobody needs to operate it. We will create spaces where a robot will come in and say, "Okay, let me control the systems in your house," and it'll connect to the MCP server that we launched, and that would be kind of the end of that. But I don't think we'll be responsible for that machinery. Also, the company that makes the first mass adoption robot will never seize control of selling them, because they will basically win the market completely, because there will be robots that make other robots. They're going to make the most robots. It's like the first company to get into AGI wins everything. Wins over humanity, or maybe AI wins over humanity. It's the same thing. The first real company that comes out with a very capable robot will be the only company left. Nobody else will make-
Ron:
I don't know. We'll have to see. In the US, there are about 10 to 15 humanoid robot companies. In China, there are over 300. I'm not pretending I'm an expert on China, but all of the IP is shared with the state and then redistributed to all the companies. So all of the inventions happening at all the different shops are being shared, and China is accelerating at lightning speed. But the US has capitalism and the spirit of entrepreneurship, and there's a lot of-
Sergey:
I just feel like if there's one company that can actually make one that is capable... Everybody makes the thing that runs and washes one dish an hour, or whatever they do right now. But there's no one company that actually has a useful thing, 100% versatile. Give it anything, it will shovel a hole in the backyard, wash your dishes, put your kid to sleep.
Ron:
I think it's going to be here in the next one to two years. If you're following Figure and what they're doing with their robot and some of the breakthrough tests they just had in the last 30 days, and then Tesla's bot, there are some really neat advancements happening. I think there are going to be hundreds of millions of these humanoid robots in society in the next 10 years.
Sergey:
I know, but it will be one company.
Ron:
All right. I think I would take a bet with you on that one. I don't know that any one company can produce the assembly line to meet the demand. I think there will be an insatiable demand for these bots, and for that reason alone I think it will be distributed in some way. But we're predicting the future, so we both could be right and both could be wrong. It really doesn't matter, although it will be fun to go back and watch this video one day in the future and go, "Hmm, I wonder which one of us was right or wrong, Sergey or Ron."
Sergey:
I'm glad that we're recording this.
Ron:
Yeah, this will be fun. All right. Let's get back to business. People didn't come to listen to us talk about robots and these predictions of the future. Let's circle back to your business. We're in the last quarter of '26, and your sales are doing really well, knock on wood. You're seeing nice growth over last year. What are some operational processes in your company? It could be meeting structure, cadences, mechanisms of communication. What's something that you guys do that you think works, that's really helping you run a better company today than maybe the version of your company three years ago or six years ago?
Sergey:
We realize as we get busier, we need to slow down more in order to output quality. If you run at the speed we typically would run, when you're busy you start circumventing some of the steps that you've developed in order to keep quality up. So slowing down is really important. We are now consciously trying to slow down a bit with everything. With the scheduling, I know it's very difficult to juggle a schedule that you don't set on site, but we request money early in order to build things in-house earlier, in order to be ready earlier, so we can pace out the necessity of spending hours on a very specific project. One of these things is that we have now almost switched completely to a 4/10 schedule. Not really 4/10, but mostly we work only four days a week. On Friday we all get together and synchronize on what was done over the week and what we're planning to do next week. If there are any holes in the schedule, we plug them in. If there are any holes in production, we plug them in. If there are any problems that arrived out of the field, we solve all of that. That's been super productive. We basically spend Friday morning doing that.
Ron:
Does that meeting have a name?
Sergey:
Yes. I call it the company sync, but I think we have an official name for it, which is not very glamorous.
Ron:
No. Okay. When you say that, is it an all hands, everybody in the company, or is it just your technicians or PMs, or who's in that meeting?
Sergey:
It's essentially all the field employees and the admin, the scheduling admin. They get together, and guys bring problems, things that don't function. I get yelled at for three hours or so. Out of three, I get yelled at for at least an hour every Friday for trying to mess with the process.
Ron:
Is it actual yelling, or are you saying that facetiously? Your team doesn't yell at you, do they?
Sergey:
No, no, no, but it's a bunch of Russian-speaking guys. We have one salesperson, his name is Joe, who is as American as you can make it. He sits in there for the first hour and he just doesn't have... He started to record this on Plaud now, and you should look at the translation of the things that... But the decibels in the conference room go up very high.
Ron:
Up and up and up as the meeting gets longer?
Sergey:
Yeah, for about an hour everybody vents, and then it's, "Okay, how do we do this?" Which is amazing, because I look at it as our team growing. They're growing in how they perceive their responsibility for delivering good service. They care about the company. We managed to figure out how to set incentives correctly. When it gets loud in the room, everyone is basically trying to get the company to operate better, right? So we make decisions and...
Ron:
We appreciate the passion.
Sergey:
Yeah, we appreciate the passion. And this coming Friday, which is in two days, we will have a quarterly meeting where we actually go offsite and everybody is present, and remote people are on call. We don't talk about work. We just focus on how to actually run the business better. It's sort of like a nugget from EOS that we implemented in 2018, so we've been doing this for eight years at this point, and it's been transformative to the company. We've worked out so many processes with it, and we keep doing it. There were years when we would skip one or two, and that was bad for culture. So we don't skip them anymore.
Ron:
Do you self-facilitate those quarterly meetings? Do you run the meeting, or who sets the agenda and runs it?
Sergey:
Unfortunately, it's been me the last maybe four quarters.
Ron:
Okay.
Sergey:
But I try to bring others in. The way it goes is we go for the culture, and I do the culture values and the reason why people work with us, in order to remind the guys what we do. We have customer interaction training, so I do that. How do you talk to the clients, and so on. We have site interactions, like what you do on site, how you handle tools, who you have to talk to. That takes a bit of time to go through. Then we have a time when we open it to everybody, and everyone voices their concerns, comments, and improvements. When we have a little bit more time, we actually have people prepare something. This Friday, we actually said that the last meeting did not go amazing. I spent too much time talking. So this meeting, everybody's going to go first, and I'm going to go last. Then we'll see what happens in this format.
Ron:
I did that recently. I'll share my experience with you. I joined one of my teams' L10, their weekly meeting. I went in telling the leader of that team I was going to be a good listener, and then I had the meeting and I was not a good listener. I was too busy talking. As soon as the meeting was over, I had instant regret. I did the opposite of what I had just committed to myself and to them to do. I had a meeting today actually with that leader, and she's inviting me back, and she's like, "But this time we're going to flip it." She and I talked through it and laughed, and we're going to make sure I'm hearing all of the feedback around the subject matter that we're going to talk about. Because it is important for a leader to listen. Listen first, act second. But even those of us who have been doing this a long time still need practice and coaching and reinforcement. It's work.
Sergey:
It's sort of like raising a child, and I don't mean to assume that I'm good at it. You have to let them make mistakes. As long as you feel like, if it goes the wrong way, you can always repair it, it's always better, because they go through it, and they learn, and they do better next time. That's how we learned, right? But we had all the responsibility. So now you're just passing that responsibility forward and praying that it's not going to go sideways. But the team is exceptional. This is the best crew we've ever had. They're responsible, and they care, and I'm happy to listen to 90 decibel screams at me.
Ron:
I would pay money just to be a fly on the wall on one of those Fridays. Your sales guy has to record that if nothing else and put it up on Instagram as a reel or something. That would be funny. Sergey, it has been a pleasure, sir, having you on Automation Unplugged. I want to thank you for coming on and being so open and transparent, and even allowing us to talk about some fun futuristic topics that we'll have to replay in a few years and see how wrong we were or how right we were. For those who are tuned in, they're either listening to us via their podcast, or they're watching us on YouTube. Where can we send people who want to get in touch with you, sir? I'll put some banners on the screen to support whatever you say.
Sergey:
You can email me, call me, or find me on LinkedIn.
Ron:
Let's go one by one. And remember, people are listening, so they may not see the screen. Where do you want to send them for email?
Sergey:
Email is Serge, S-E-R-G-E, at cravtsman.com. It's spelled like the word craftsman, only with a letter V. Cravtsman.com is our website. Cravtsman_inc is our Instagram, and I have no idea how you can find me on LinkedIn, but I'm there.
Ron:
But he's there. Just search for Serge. And then your phone number. Is this an office phone, or is this a cell?
Sergey:
This is my direct line.
Ron:
Oh boy. Watch out, world. Let's call him and ask him if his refrigerator's running. No, I'm just kidding. Do not call Serge and ask him if his refrigerator's running. His phone number is 201-400-6443. 201-400-6443. Sergey, thank you for joining us on Automation Unplugged. It's been a pleasure catching up.
Sergey:
Likewise. I appreciate you guys having me here. Thank you.
Ron Callis is the CEO of One Firefly, LLC, a digital marketing agency based out of South Florida and creator of Automation Unplugged. Founded in 2007, One Firefly has quickly became the leading marketing firm specializing in the integrated technology and security space. The One Firefly team work hard to create innovative solutions to help Integrators boost their online presence, such as the elite website solution, Mercury Pro.
Resources and links from the interview:
Show Notes:
Links: