Leap to Scale
Leap to Scale is for technology curious leaders of service firms who want higher margins without adding headcount. Each week, Justin Davis and Greg Ross-Munro show how to turn firm expertise into repeatable, sellable technology products: SaaS, packaged workflows, and AI-powered tools clients can buy again and again. With AI, more of your know-how can be captured, standardized, and protected as IP instead of being rebuilt in every engagement.
We focus on practical decisions: what to productize, how to price it, when to build vs. buy, and how to use AI responsibly without risking delivery quality or margin. Clear steps, real tradeoffs, usable examples.
Leap to Scale
How AI Agents are Automating Business Growth
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
In this episode of Leap to Scale, Justin Davis and Greg Ross-Munro sit down with Doug Smith, Director of AI Operations at Element 451, to explore how AI is transforming the way service businesses operate, automate, and scale. The conversation covers everything from hidden operational data and AI-driven workflows to rapid prototyping with tools like OpenClaw and AI agents. Doug shares practical advice for identifying automation opportunities, avoiding “AI slop,” and using data exhaust to uncover growth opportunities hiding inside everyday business systems. It’s a candid and highly tactical discussion about where AI is genuinely creating leverage today and what firms need to do to stay competitive as the technology evolves.
All right. Today we're meet talking with my friend, Mr. Doug Smith. Our friend, Justin, Mr. Doug Smith. Yes. Doug and I actually graduated from uh together from the same Master of Science in Entrepreneurship and Applied Technologies. Was that the degree that we got? I think that's what it was. Yeah, it looks like it's a mouthful. Back in the day, me, I would think Doug has like has always stood out as somebody's always thinking like that next layer deeper than everybody else, especially me. This is like from everything from politics to the crypto conversations I have with him. But never in like that growth hacker, fluffy BS kind of way. And I think Justin, like you and I, he's he's always somebody who like never really seems to just accept how things work, and there's got to be like a better way. Um officially he's the director of AI operations at Element 451, where I think he focuses most of his time on putting AI into real kind of ops systems. But what makes Doug interesting for this conversation and and and our audience is I think partly the path that he took to get here, he started in I think finance and strategy, and then he moved to operations improvement, and then like got into data science and AI consulting, and now he does like AI ops. Um and so I think he's gonna bring like a super practical viewpoint to all the things that we talk about all the time. He is also somebody who's like hugely involved in the tech ecosystem uh here in the Tampa Bay area. He like he shows up, he participates, he's in the signal chats that I'm in, um, always sends me like a million different things every day that I've got to go and read in white papers. Uh, and I think while he is not a programmer at the core, he's still a type of hacker. Um, not in the security sense, but like the classic sense of somebody wants to like tinker with systems and things. So uh thoughtful, curious, and like just one of those people I really like. And so, Doug, thank you very much for coming on and chatting to us in our audience. We really appreciate you, man. I appreciate it. I always enjoy chatting with you all whenever I get the opportunity to do it. So yeah, take it seasons in camp. Normally we do it with like beers or we're like we're yelling at something, but like this time we got a microphone in front of us and we'll do it a little more ready. All right. So uh, Mr. Smith. I hope you're doing well today. Tell us a little bit about your origin story. I think I touched on some of it, but how did you get to being where you are today?
SPEAKER_02Yeah, so I worked in the the world of finance and like background, went to school and economics, did a bunch of like trading and stuff like that when I was in school and after, and ended up in a pretty cool like position working for a venture capital firm, looking at deals, looking at new ideas, and just like that sort of like the newness of the world. Like, you know, this you know, the the world is going to be a better place because people were working on solving these very thorny problems to to save time, to save lives, to save resources, etc. Um, then I went on a path of worked with a private equity group that was in that pursuit in the education space, and then understanding, you know, you start going to this like, oh great, this is the numbers, the one zeros, how these things make money. Then you get the next layer of like, well, how do they stay open and how do they actually solve and solve and remediate the problems of like students and sort of digging into those problems, like the systematic challenges, take Maslow's hierarchy of needs. How do you figure that out? Well, data, right? So learn data, figure out how that can be applied. And you know, I had an econ background, but just it was, I would say, one of those things you didn't really use until you're several years out of it. You start looking at regressions and building regression models, and you start figuring out these are the the sort of signals of why a student might not be successful. Then you say, like, great, we figured out who's not going to be successful, what can we do to remediate that? Then you start working on like the systems and processes and change management, the people processes, the technology to remediate those things. And that was sort of where I've gone. And uh to me, I always love this like super ambiguous problem that was never able to be solved before. People have all tried it. I'm like, there's a way to probably do it if we just think about it in a different way. Um, I'm a big proponent of like taking other approaches that other industries have taken to solve the exact same problem and seeing can maybe apply to a more traditional business. So that's sort of you know, reclaim and demonstrate. Don't do hardcore RD. Like somebody has probably figured out this problem for you elsewhere in a different use case. And I mean that's the beauty of academia, too. Bountiful amount of research papers that have probably told you how to solve for the set of problems you have. And now you just amplify that with LLMs and you just get just hooked. And I will say this much in uh Greg and I you you say it often, everyone says this often. If you turn off from AI for a week, you're like an entire year just flew by of new developments and new changes. It's just uh it's just fantastic to see, Frank. I I'll say like it's one of the most it it is I'm like polar opposites, most super excited about it, but know it is very much going to disrupt a lot of human people's lives at the same time. Like people have been able to accomplish and do things that they with was thought impossible or take hundreds of thousands of dollars or millions of dollars or tons of staff to do, and they're doing it themselves with agents now, to well, do we hire those things? Do we ever solve those problems again? Do we need people to solve those big complex problems, or we just keep doing more? I don't know. But uh that's my take on it.
SPEAKER_00That's sort of how I got to where I'm at. So how did you how did you get from like the the PE and um uh the PE side or the like the finance side into the education side?
SPEAKER_02Yeah, so so I went into working out with the portfolio company um and stuck stuck stuck it out there and then went into the world of consulting.
SPEAKER_00Um they put you in like an ops role, they put like the PE guys put you in an ops role.
SPEAKER_02Yeah.
SPEAKER_00They were like, go fix this. Okay.
SPEAKER_02And that's I mean, that's what often I was was like the fixer. Like you you can look at things in a different way and help these people that are really competent. Like they're they have a ton of domain experience. They don't know how to translate domain experience into technology, into data to be able to act on it and to do it in a scalable fashion, right? They can probably brute force it by you know, back in the day pulling all the spreadsheets together and spinning out countless hours and figuring it out to like that whole access database or series of spreadsheets be put into a SQL database that then can be run with cron jobs and then that can produce the report versus you having five managers spending their entire day building that. Yeah, that kind of stuff.
SPEAKER_00So I guess I guess I guess I mean this br brings up like a really interesting point for like that we run into a lot of times, which is that people come in to talk to us and you know they they understand there's like a huge amount of probably hidden value in the the data that they have in their business. Um but they like a lot of companies we find like they don't think they have a ton of data, but they actually have like a lot of operational signals that you know the the this stuff cut this this inputs coming in all over.
SPEAKER_01Yeah, kind of data exhaust, if you will.
SPEAKER_02Yeah, yeah. There's like little artifacts of data like sprinkled around, and sometimes there's like some things that need to be done to like massage to make it all come together. So, like, I mean, all companies have this, right? You probably have a series of clients, it could be like a handful to a ton. Is there anything that could be like signaling why they might be leaving you? Like, do you have a product that like feeds the data out? Do you have if you don't have a product, do you have something else that could be like a key signal? Like, yeah, your accounts receivables, they stop paying you. That's normally a signal. Um, if it's like a restaurant, even right, like you have the same people coming in the door and they stop coming in. Like, why did they stop coming in? What was the root cause issue with this? Um like, and all often it's just like you start having the conversations and then you're like, I could probably create a way to standardize and collect the information. Yeah, great. Now I'm standardized and collecting the information, so I'm creating a layer of quality data. And I think that's the biggest challenge that everyone has in some capacity, from huge mega multi-trillion dollar value companies down to um the mom and pop shop is like think of how you want to collect the data in such a way that you can look at how, what, and why that thing happened and very clearly some type of unique key. If that is a if it has to be as rudimentary as an email address or a telephone number or a physical address, those those can work. Best thing is to probably get something else. Um figure out how you could, and if need be, create that unique key, figure out how to get into your multiple systems. So you've got Greg, Justin, Doug as customers, give that unique key, put that into here's your your QuickBooks to your inventory system, to your product solution, to whatever else it is, um, and start associating that work. Um are they Facebook followers of you? You can start doing all these things of like, oh, that's Doug Smith, same Doug Smith that's my client, put that unique key on him, he's liking me, that kind of stuff. And also you realize they're influencers. Like, there's a whole ton of hacky things that you can do. So going back to the hacky thing, you do these hacky things, but they'll get you to a place that you were before of like giving you insights. Um I'll say the biggest challenge of any organization and every organization I've gone into is there's a often though, a hyperbelief that the data is very good and there's a ton of it, but no, do people actually believe that?
SPEAKER_00Yeah, they do.
SPEAKER_02Yeah, like they're very much like we have great data architecture, great data stream, and you walk in and like can't answer basic questions of like, how many people are doing and utilizing these types of products in the future? Oh, we can figure it out from mixed panel and like go over here and spend like five hours creating this like one-off SQL thing. And I'm like, well, we have like tons of products, like that's gonna be really hard. If I'm an account manager having to understand, like, how are they using the product? They're not like experts in data, they're not experts in querying this. So, like, they're gonna have to go click on the instance and figure it all out and spin that up. So that's a hard thing to do. And that that's that's the beauty of a lot of what's happening, though, is if you are able to gather the data and position it in terms of what is the tacit knowledge often as like as founder, director, or manager, or whatever it is going to be, you probably have a tacit knowledge of like how a thing is supposed to be done. How did you figure that out? And can you replicate that with data to then translate that and give that over to a staff member, like on a consistent fashion? Greg, I imagine you have a very different approach to solving a very angry client than a lot of your staff do. Just how, but how do you codify that and how do you translate that? That's a hard thing to do.
SPEAKER_00Um probably I try and make them laugh, to be honest. That's generally what I try and do. And then once they're once they're chuckling, then they get less angry with you.
SPEAKER_02So like but go you go into this, like there's a series of things that you can do. Okay, you don't have to laugh at those.
SPEAKER_01So I think what what's interesting about this to me is that like because I think you're right, like there's kind of like a few things that are true. A, a lot of organizations sort of do have this often my naive belief that like they have a lot of data and it's good and it's organized well, right? And there's also the converse is also true where there are a lot of companies that actually don't realize that they actually have data, right? That like they don't realize that, oh, my invoicing system has lots of data in it that I just never thought about that it's sitting back there. And and and what's interesting is that before now, because to use that data wasn't really something anybody, a lot of people paid attention to because to use that data meant you had to know how to get it, which meant you had to know how to get to it, SQL and integrations and all that APIs, the APIs and all of this, you had to know how to to analyze it, which means you had to have like a real person that knew how to like do regressions and k-means and all of this kind of stuff, right? But now you have that. And so all of a sudden, this sort of like exhaust that was mostly useless because it was not something you could do anything with, now all of a sudden has instantly turned into this extraordinarily valuable asset that every single company in the world is sitting on top of. Which actually makes me think then that there would there might actually be a wave of sort of data governance work to be done as consultants and products and all of that to basically say, well, we just turned a bunch of useless garbage into gold that you're now sitting on, but somebody still has to help you refine it. And that's probably a layer of service that's coming in the market, I would and an opportunity, I would assume. Doug used to do that job.
SPEAKER_02Yeah. So that's yes, that was and then at the end of the day, it's way easier to do now. But like useful example, right? Who are your clients? You have you know who they are. You probably have a job title, you probably have their name, and you probably have their company. You could go to something like Clay and say, like, hey, find me lookalikes in this geographic area for a new business. Here's my ICP, so my ideal customer profile, here's that list. Here are people that are just like that, have similar job titles and names. Boom, like that. Okay. And then you guess what? You find their LinkedIn, you find their email. You like so you can very quickly like sort of scale yourself up. Um, one of the real novel ones that I heard though, um, and Greg, it might have been from you or it might have been somebody else, but like these individuals that do like trades type work of like they have a shared client base often. Like the person that has to have their lawnship done. This is for me.
SPEAKER_00This is why we keep talking about lawn care, just like talking about lawn care all the time.
SPEAKER_02Yeah. It's just like that that is the pathway, it's like the pathway in of like your your law, your lawn needs to get taken care of. Now, if you have a very astute, and I think a lot of these people are entrepreneurial, like, yeah, they're like, I have a skill set that like I can also paint stuff, I can also do like white carpentry, I can fix the fence, I can do hardscaping, or I can do the window trim, or I can, you know, I'll charge you an extra 25 bucks and I'll mop all your hard surfaces outside so they look great. And you're like, what an upsell. What an upsell. But um land and expand. Land and expand. And the thing is, you have everyone's addresses and you're like, hey, guess what? I can go on Google Maps. I don't have to drive there. I can look at what they have. And I'm like, hey, we can estimate what this might be for you.
SPEAKER_01Like and I tell you what, like this stuff nowadays. We were having this discussion earlier. There is like this sort of dislocation in the market right now. We're recording this in the spring of 2026. And at right now, the state of this is that there is an insane amount of power available out there for nearly free that most people don't know about and aren't really using, which means that if you lean into it, and if you're listening to this podcast, you probably are, um, and invest in that for a while, you're gonna have a real sizable advantage until that market dislocation kind of heals itself back up, which it will, but it will take some time. But then you'll be ahead of anybody anyway. Right.
SPEAKER_02You'll be you'll already be ahead at that point. You're you're gonna be a survivor in survivor in the tsunami and/or when the tide goes out situation. I mean, that's that's that's the reality of it. It's like the ability to maintain and be nimble in what you do. Um, and I think that's that's one of the core things that I that I see here. Like I've you know, transition over to like the open claw sort of emergence of things, and like people have really gotten claw pilled, and I can totally understand why.
SPEAKER_00Like, so you might need to Doug, you might need to step back and explain what open claw is. Yes.
SPEAKER_02All right.
SPEAKER_00So open claw is without using the word lobster.
SPEAKER_02Yeah, without using the word lobster. So it is an open source framework that allows for individuals to download that onto a computer device, a virtual machine, which is a computer in the cloud, and have access to it. And you plug in an LLM, large language model, your chat GPTs, your quads, your Gemini's of the world with the fancy API key. And you can have conversation and then can build a lot for you. When I say build a lot for you, different types of web web apps, web technology, task processes, and things of that nature. So all that sort of wonderful creative idea of like, what if I could do that today kind of thing? You can have a conversation with OpenClaw and it will try its best to do it. And a lot of people have had a lot of success with it because the large language models themselves have become and gotten to a place that they're performant enough to actually execute against that. Will it be the most 100% efficient way? Probably not. Will it be hacky? Yes. Will it potentially be a security nightmare? Yes. Would I recommend you downloading this on your personal your personal computer? Questionable. Could you get a digital?
SPEAKER_00No, I'm I'm gonna say no, unless you really know what you're doing.
SPEAKER_02Yeah. Do not install it on your personal computer or your corporate computer. These are these are big no-nos right now.
SPEAKER_01Um what kind of computer then? A library computer? What other kind of computers are there?
SPEAKER_00Well, Justin Davis, let me show you this small little Mac mini that I have sitting on my desk. I mean literally, which has nothing. This is on a separate network to the office. This is not even on the same network that we're on the Wi-Fi system here.
SPEAKER_01So I'm a little more cowboy because mine is hooked up to my home network and has access to all the devices on my network and all of my smart devices and has carte blanche access to everything. So uh I'm a little bit more.
SPEAKER_00You are a you are a dive without testing the water kind of guy. I know. Generally speaking. Yeah. I'm my my plan for this little Mac Mini over here is to have another Mac Mini running open claw on it as well. But that one's hooked up to a shotgun. And the second that the other one steps out of line, it just pulls the trigger. That's the plan.
SPEAKER_02You're just you're just gonna get like the notification that there was a loud noise at the office one day 2.30 at night. You'll see you'll see that you'll see the logs of what happened.
SPEAKER_00Yeah, I the the expression I don't trust it as I wouldn't trust it as far as I can throw it is really um inaccurate here because this is a Mac Mini. I could throw it a long way. Long way, right.
SPEAKER_02You can yeet that thing across the pond. So, and not to like just side skirt the the potential security risk. And th those are very valid, you can find a ton of web content around that. The unlock, though, for people, I think has been pretty amazing. Like I have friends that do commercial development. Very straightforward. How would you figure out where to go do deals, how to figure out these kind of things, like what available parcels are out there? And these are really intelligent guys, girls that are able to do this, which is like they know how to go do that, and it's really hard and it's costly to do it. And now what they found out is like they can go to open call, they can ask it, like, hey, can you build this thing for me, this prototype? And it and it works. And they're like, oh wow. So I'm listening to all the board board of you know, board of planning for every single county um in the state or in my target state. Or I was able to map utilizing um uh a GIS software package, all of the potential available properties that would be for sale on a particular interstate in an area that I'm trying to develop on. That work would have been extremely expensive, very time and toil. But the benefit of this is now like they figured this out and they're like, I probably want to, and uh every time that I talk with one of these folks, they're like, how can I turn this into a product and sell to other people? And I'm like, so there you go. Like now you have to harden it. Now you have to like you go through the sort of traditional SaaS-based process of like, it needs to be secure, you have to be able to stand it up, it's gotta have all these other critiques. But there's value there. You validated the concept and the idea around it. Um so that I think that's the beauty of what's happening with open call, with replit, with lovable, is like the concept idea can gain and garner traction to validate something relatively quickly. And then you can then explore working with folks like the level of folks at source code to turn that into a robust, repeatable, scalable solution. Um and you've in part curtailed some of the the market risk of like, is there does this solve a pain point for people? Um and I and I look at it that way. So I I freaking love it.
SPEAKER_00So if you're if you are um if you are like a service business or anything, and like you, you okay, you you've you've got some data, or maybe you're like working on collecting it, like what is the first where do you start? What's like the first automation or the first like system thing that you build? Like how do you how do you identify what the low-hanging fruit is or or what what you should do first?
SPEAKER_02I think it's too I I think you start with dollars, right? Dollars and cents. What's what's gonna bring in net new dollars and then what's going to make sure you get paid on something, right?
SPEAKER_00Like understanding, like very simply, hey, here is so not you're so not cost reduction. You're not looking at like No. Okay.
SPEAKER_02And the reason why is like that that's on a val that's gonna help you validate your data system of like what do I need to be able to do to bring. People all the way through. Um, and like where are the cost centers at?
SPEAKER_00Like, can you can you say tell tell tell us more about that? Like, what do you mean, like where the dollars are?
SPEAKER_02So, what is bringing in dollars for you? So how at the end of the day, how did the customer end up with you in some capacity? Like you probably might have a marketing person, you might have an agency that's doing stuff for you. Have you ever asked to like pull all that data together, have them pull the data for you, send it over to you, pass that into an LLM. Ideally, you have a business account that's secured, you know, it's not passing it into the ether. We're just gonna gloss over that. But you start getting answers out of like where potentially people are coming from you. And then like now you start realizing where am I spending my dollars and time on gathering and bringing in front front office people? Or do I have like sales reps that are or are not doing types of work? I've got a buddy who's utilizing uh, you know, an open call LLM like solution to review all the calls and transcripts that the staff that are having of like, what are the questions that they're asking? What thematically is happening? Like, what is happening on this situation? Because that would have been absolutely unscalable for a human to do. Like, there's no way that you're gonna probably listen to 10 to 40 hours of telephone calls every single week for all of your folks. You would have to hire one to two full-time people to go do that for you. Now with an LLM, you can. You can start thematically understanding, like, hey, this this client's angry at me. Okay, great. Here's my summary of like why they're angry at me. Like implementing those basic solutions, super useful. The summarization of the interactions you're having with folks is super helpful.
SPEAKER_00Your suggest your your suggestion is like start at the top end of the funnel rather than like a business process that you want to automate. Interesting. Like I I I've actually like for a long time thought the opposite is true, which is like trying to find out what what's like burning the most amount. I'm not saying you're wrong. I'm just saying like for for me, it's like what's burning the most amount of time where those people could be doing other things.
SPEAKER_02The time and toil. Well, there's a component of this, right? Which is like when you start going through that, you start realizing where the time and toil is at. So to help you basically identify where you would start getting the the time efficiencies. So let's let's take this like the traditional sales role. Where is most time and toil done? Prospect research and finding it, and then documenting that into the solution. What is the biggest time suck? The biggest time suck is documenting the outcome. Can you automate the outcome of populating the information in the Salesforce HubSpot notion? From like the conversation. From the conversation, right? So like that that's just time and toil that you yourself or somebody else is having to go do. It's like writing meeting notes. Exactly. Like the outcome. What was the outcome of this attempted thing? Now, once you start translating that, of like you then you figure again, start you keep edging on like, well, what is the next thing for me to look at from that place of like, okay, here are is now I start building you at the end of the day, you also start building a data pipeline of like, here's the conversation, how has this person become a client? This is what's important to them. Now I have a better picture. Now, if you have account management or you're doing account management for your clients, you know what actually, why, why and how, where they came from to get to you. And you now have some actionability there too, on like how they use it, what their aspirations are, like how the potential you work with them. And then as you continue all the way down, you start figuring out when you start talking with yourself or other folks, then you start seeing like where are people like the massive time sucks. Like, and there's always tons of tons and tons of massive time sucks. Um, very basic ones is like the account prep research for an account manager, account prep research for a salesperson. Um, all those things go into it. If you've got um, you know, let's say you're owned by another organization, you have a formalized board, like the drafting, the creation of the quarterway business reviews, the monthly business updates, those are all really massive time sucks for entrepreneurs and founders, etc. Is there a way you can automate that away from you? And you know it's gonna come every single month, it's never not going to be there. Um, but I do give a word of caution of like, don't automate the thing that takes like one or two hours to a month to do. Like, do the thing that is like one to two hours a day, and or it's five to ten hours a day across the entire team. Is that is there a way to solve for those kind of problems? Um and I characterize that in if you go into the world of like the four quadrants of like what's important, high impact, low effort, you start figuring out those kind of things. Um I think one of the logical ways to, if you're sort of like, what should I start with, go into these LLMs, Claude, ChatGPT, have an interview about your business as like a business coach expert, and like grill you on those type, those kind of things.
SPEAKER_00That's interesting. So what like you would um you would say open up ChatGPT or Claude or whatever, and then what would the what what's the question you would ask it?
SPEAKER_02What's the context to say I'm John Doe and I run a hardscaping business in the Florida market?
SPEAKER_00We're ready. Every podcast is gonna be about hardscaping from now on.
SPEAKER_02I'm fascinated by so uh and I have X amount of employees, I make X amount of money, I want to grow. Here is what my business is doing so far. Help me understand what I should, what I should know be thinking about my business to grow it. If that's personal, personnel management, human capital management, to areas that I should be investing in, to data that I should be looking at. Like what do we just go through each one of those pieces? Because that's the magic of these LLMs, is that you really do have some of the best and brightest ideas encapsulated in these things. You just need to know how to ask it in the context of who you are that can make it actionable for you. Um, and I think that's that's the a key thing as you work with LLMs, giving it context of who you are and what it's working with is so important. And it actually is the foundation of everything we've talked about today. Is like the data that goes into the LLM, the better the answer is going to be for you. The more likely it's able to answer and solve the thing for you. If you're going in and developing and creating a new AI agent solution for you, literally do an export of your key reports that you have, give those as context to the agent so it knows, hey, you have these fields with this type of data. It can do a data audit for you of like, where are you missing information? Where is it not right? You've got data that is not actionable because the way in which it's formatted, I can automate and format that for you. And now you have clean telephone numbers and clean emails, um, all that wonderful stuff you start being able to get at. So um, that's the sort of interview process that I that I think about here of like, what do I um and I'll share this with y'all is like there's one of these that I built, which was um founder mode. It was that um off of Paul Graham, Paul Graham's commentary of founder mode. Took founder mode, threw that into a GPT, and say, I need you to act like founder mode, talk to me about myself and what I'm doing. So an action plan of one thing I need to do every single day, things I need to be asking of my employees, things I need to be asking of like people around me, things I need to be asking of my vendors, like this sort of like coach pushing you forward. Um, that's one clear use case of LLMs um to scale it. And then obviously you can do a ton with the agents, and I go on for hours about this. Obviously, you can see I'm passionate about it.
SPEAKER_00So so you're suggesting everybody goes to Paul Graham's blog, which hasn't changed its style sheets since like 19 no, uh 1999, maybe?
SPEAKER_01I don't think which I think is like a low-key like credibility.
SPEAKER_00Yeah, no, I mean indicator on the web now. It is cool. Uh um, but it does look like it's from a from it is from like two decades ago. Um and the founder mode thing was, I think, I don't know, two or three years old that that post, but um, go find it, copy it into your LLM, and then say, do all these things and ask me these questions. And then one thing that I always uh advocate for is off after you've been interviewed by the founder mode thing, save that output somewhere, save it into a text file because you might that stuff is important enough where those answers are important enough where you might want to shift LLMs or you might want to use it for something else, store it in a mockdown file, store it in a text file, save it onto your desktop.
SPEAKER_02Yeah.
SPEAKER_00No, it's all about context, all about this weekend. Yeah, yeah. I have a GMT if you want to, it's already there. Okay, yeah, I'll share it with you. Send it out. We'll we'll put it in the show notes.
SPEAKER_01Um, you know, I kind of like the idea also of maybe even taking that interview concept like that and using it a slightly different way. You made me think about this, which is it it's interesting to think about going to your internal employees and either interviewing them about their job, recording it, and then like just saying, like, talk to me about what you do, what's the worst part of your day? What are the things that you do? What are hey, what are your tricks? What are the things that you know, capturing tribal knowledge through interviews and then having LLMs parse it and then create process um ideas out of it is probably a sneaky, high-fidelity way to generate a lot of ideas very quickly that you could probably do in an afternoon over lunch.
SPEAKER_00Yep. Yep. Um I'll I'll ask you one last quick um kind of question, which is a little bit more future focused in, which is that if if you were like if you were the CEO of a service company, what would you do to like design kind of the ideal operations dashboard? And maybe I'll change that question up a little bit to say like what would it show now? What would it show in the future, in your opinion, that most businesses aren't tracking yet? Like, what am I not tracking or what is what are people not tracking in your hardscaping business that that they that is gonna show up on their like when when when the dominant hardscaping company two years from now is like crushing everybody else, what are they tracking in their their CEO dashboard that people aren't tracking today?
SPEAKER_02Probably the human-to-agent interaction in terms of success successful resolution and positive outcome and recommendations by the AI agent adopted by the organization, which is the agent itself, because of in theory, the amount of information that it is collecting should be able to help provide novel things. So you'll basically have human developed solution versus AI-based generated solution. Also, you you probably will have this at some point in time. We sort of track it right now, which is like the ROI for how much like element clients get back to be people, the human literally it's just gonna be a gauge of human unlock, which is the amount of time your staff didn't have to interface with non-productive work for them for what they do.
SPEAKER_00You're almost talking about like feedback, like feedback loop stuff, right?
SPEAKER_02Like that's that's what you become. I'm I think every CEO, every sort of manager of things, you're gonna be an orchestrator of agents and feedback loops in A in AI workflows in some capacity. Yep. Um you're still gonna be interacting. If you're in the service industry, you're gonna be interacting with humans just straight up, unless you have a like drone swarms flying around cutting down trees or something like that. Great business concept. If you thought about it, you should get into that. Um there's gonna be a dirt, there's gonna be dirt, there's gonna be a dearth of drone productivity needs in the country, I believe.
SPEAKER_00So somebody's still gonna somebody's still gotta like convince another human being to buy my drone fleet rather than others, and that's gonna require people, unfortunately.
SPEAKER_02Everything, and I I say this now, it's just like I think the the incumbent noise on text messages, emails, LinkedIn, etc., you're getting drowned out. And the nature of sales is gonna be more business. It's gonna be more human. I think we're actually again sort of the tide went one way, I think it's gonna go back the other way.
SPEAKER_01Almost it would like choke itself out to the point where it's almost like it kills the host, right? And like we're gonna go back to just going out and having beers with people because like it's just gotten too noisy in the other space.
SPEAKER_02Yeah, and like if you are say relying on like your agents to do research for you of like who should who should I be working with, like to be honest, the social proofing of Google reviews becomes even more important than it ever was. Better business better business bureau becomes even more important than it was because guess what? Like those are artifacts and signals that people were basically going to highlight against. Um, so I it's it's gonna be a I I'm very excited for the future, to be very frank. Like, um, I'm super excited for folks that especially in the service and trades industry, of like if English is your second language, of like great, now you have this high fidelity, like Betty Sue sounding tell teleanswering person for you who who can answer the phone for you, you sound great, whatever, you show up and you do your thing, they have no idea that you can barely speak a particular type of language. That's gonna be that's awesome. Like, that's an accessibility thing. Um, or conversely, like if you're hyper-introverted and you hate doing all those kind of things, then you have agents calling on your behalf to go do do some of that work. Um, and also like it also can make you do the work that you don't want to do of just like, hey, just go solve this problem for me. Um, go interface with this vendor, make sure I get paid. Like, pester them until I get paid.
SPEAKER_00So well, I really want to make a joke now about how this stuff is gonna open doors, but I think we've already like uh we've already beaten that one to death. So well Doug, I to say thank you so much, man, for taking your like for spending time with us today. It's I've I think we've really enjoyed this. I uh there's things I've learned that I'm gonna go and do. I'm gonna go do the um please send me the uh the founder mode uh GPT, we'll we'll share it as well. Um and I I just want to say like I've always like I love the way you think. You're one of those people who like you look at things in a uh and a and problems from a like an angle that I appreciate. And every time I every time we talk, I feel like I walk away with new ideas, and obviously not to mention like the 10 new AI tools you send me on signal every week that like I cannot keep up with.
SPEAKER_03But I re I really appreciate it.
SPEAKER_00That's my job.
SPEAKER_01Yes, that's my job. What are the top three, drop the top three tools you would recommend to the people here?
SPEAKER_02Uh if you are like in the like office work and have to do with a bunch of data stuff right now, if you have not gotten quad, uh quad business, quad quad enterprise with cowork, you're sort of it's a you're being foolish. Um work. Yeah, like I think it is incredible for folks that need like a data analyst but don't have the the chops to do it. Um you need to have connected systems, so that's the other side of it. You need to have like a good CRM, or you have to have the ability, not even the ability, you just need to have a CRM that has data, has some degree of um pieces in it. I think it's fantastic. It can do great stuff too, just even just like crunching data and doing research and work for you. Uh tool, other other things that I just really love. ChatGPTs actually get again, I think ChatGPT is gonna come back around, guys. I think this whole open AI uh uh bringing in open call under the foundation is going to be like huge for them. Um that just yeah, to be honest, you look at what Minimax is doing, Kimmy and Kimmy's doing it, these are other LLM models. They've all basically taken Open Call and plugged those in. Open call very logically makes perfect sense to go out and explore and tinker with. To be honest with you, it's like this sort of the world that I think we all want to be living in of I ask the thing and it solves the thing for me. I have the home auto home automation thing. I have a friend who has solved what he characterizes like I spent numerous weeks and nights trying to figure out how to program my home automation system, and I solved it in 15 minutes with OpenGlaw, of like that type of stuff. Um and then the other fun ones, like check out like Vappi and like retail.ai, like the voice agent ones are really cool. Uh just it's just the nature of it of like you you have a personal assistant that can do work for you. Of like, you always think about this now, and I think this is what's so novel. Uber sort of allowed for us all to have like personal drivers, open AI and AI itself is all going to allow for us to have like personal assistance and personal coders and tech. That's a really, really great comparison, actually. It's and the thing is, it's all being subsidized by the same groups that backed Uber and all these others. So the cost to do this is dramatically cheaper than it's ever been. Um and at the end of the day, that's really it. That's really it. If you got a little kid, the Gemini Gemini um storybook is a super cool little solution. I think it's free on Gemini. You can um just tell the story that you want to have, and it'll create the imagery and has a nice, calm, soothing uh female voice. Read the whole storybook. Um you can add it.
SPEAKER_00I I built a version of that two years ago because it's like a lot of pay. The whole damn thing. Yeah, that's the bigger.
SPEAKER_02Yeah, the idea you have in your head, somebody who's probably executed against it, um, and some capacity. Yeah.
SPEAKER_00Um and better.
SPEAKER_02Yeah. Or the really dumb idea that you think that no one else would care about. And I so I'm on the board at the History Museum, and I've done this where we created a GPT, or we're working on this, created a GPT for certain like pioneers of St. Petersburg, that you have like conversations with them of like, and but it's based on who they were, based on their writing and all this other stuff. Who else would care about that? Like I did the same thing for one of my my great-great-grandfather, and who else would care about that? Like my five five five family members of being able to have a conversation with this person and what they would probably probabilistically say, again, probabilistically say, not deterministically say, um, at a period of time. So um that's the coolest aspect of it all.
SPEAKER_00And Doug, uh last last question, then is I I think I've maybe asked you this before. Is there a book that you're gonna write? Because I'd rate it, I'd buy it. Come on, what are you right? When when's your book coming out?
SPEAKER_02I have been thinking about that. I should probably just like knock that out. I should just force my open call to say you're gonna write a chapter on these things and how to think about it. Um look at the end of the day, I don't think if you're somebody that is curious and likes to learn, you've probably had there's never been a better time to be alive. Like, hands down. Like could not agree. The the capability of when I was running certed, um, like just the the things that I learned that I'd never thought of because when I was it's all service industries is like how you actually properly mop a floor and like the chemicals that you put into it. I'd never thought I would think about that, but I'm also like, how would I ever got these answers without having an LLM to be able to provide that at scale?
SPEAKER_00You could mop the floor, brother. Yeah, mop the floor. And the way in which you mop it, you don't want to mop yourself into a corner now, do you? No, no, you don't. Don't install open claw, by the way, everybody.
SPEAKER_02So let's just do not install open claw unless you're doing it on like digital ocean droplets or you want to give your information to a Chinese company. Yeah.
SPEAKER_00All right. Well, uh I'm I'm feeling unwelcome our our new overlords. So I'm going publicly on the record. I'm still waiting for the Doug Smith book. Um, I feel like there's a lot of this that needs to be written down, make its way to paper, which might be obsolete in like a week, but I still am waiting for it. Doug, thank you so much for being gracious with your time spending with us. Really appreciate you, my friend. It's good to see you. Thank you for the conversation.
SPEAKER_03I appreciate y'all. Thank you.
SPEAKER_00Thanks, Doug. Have a great weekend. Bye.