DataX and the AI Tools Built By Remodelers

On today’s episode of the Contractor Growth Network podcast, we are talking to Peter and Elliott, the owners and founders of DataX. DataX is an AI optimization company specifically for contractors. Peter and Elliott both come from the remodeling field and Peter has been on our podcast multiple times talking about his company, RanneyBlairWeidmann.

It’s a very cool grassroots start to the company and so we’re happy to have them on the show to talk. A little bit about how remodelers and contractors are using AI in their businesses today.

Welcome back to Contractor Growth Network podcast. I’m Logan Shinholser and today I have Peter Ranney and Elliott Wittstruck on and we’re talking all about probably one of the nerdier things in the construction industry right now, which is AI, but really more so how do you actually institute AI within the business? So I know Peter. Peter’s been a client of ours for years now with RanneyBlairWeidmann and he and Elliott.

Well, actually, I’ll let you guys tell your journey. Peter, Elliott, take it away. Give us backstories. Peter, you go first.

Like talk about like you are, you know, Ranney Blair Weidmann. And then basically the things that you were doing that kind of brought you up to DataX. And then Elliott will have you do the same thing because I’d love to know how like the journey together kind of came to be. Yeah, well, RanneyBlairWeidmann’s been around since 2014 and really from the start, like most businesses, you just kind of scrap and claw, do whatever you need to do.

But as we got a little more advanced, little mature, we realized we need a system. We started with the blue dinosaur BuilderTrend. It was a great system for us at the time, but ultimately transition kind of like as were again, the business was maturing. Like, okay, we want something like a little better, a little faster, more like tech driven and ultimately found JobTread.

And really what that started creating for me as a business owner of like how can I alleviate my pain points and my problems. I’m a very process driven. The consistency, like naming conventions, things like that So like zapier became really a big part of what were doing. And really truthfully, like when I signed up for job trade, I like I had no idea what API meant or that it was kind of like open, you can connect other things to it.

Like that was a whole new world for me. And so I was like, oh, what’s the zapier thing? Let me just start clicking around in here. And then lo and behold, Later I have all these apps and they’re like maxed out on actions.

And I’m not saying that’s a good thing. It’s probably my inefficiencies of like, can’t do it better. I just, you know, just brute force it. But these app ears led this idea of like, how can I create more automation in the business?

Ultimately met Elliott at Job one of the JobTread connects, I think it was Connect two possibly. And were asked to speak on budgets and cost groups, things like that And from there just kind of formed a friendship. And really this like, love of technology and automation and AI wasn’t, I guess it was a thing, you know, ChatGPT was still. It was, you know, very prevalent back then.

But there’s no way to like connect things together. And so really this idea of how do we connect our systems together, spawn DataX. And now AI like adds this whole other like, layer that is possible there. And real fast, before we jump into Elliott’s side, Peter, talk to us about like, what is Zapier and what were you using it for within JobTread that you had all these things called actions that you were setting up?

Yeah, like the two main ones I had. So Zapier essentially like allows you to connect software together. Right. Without being a coder or anything like that It’s all plain English, clicking buttons, like very intuitive kind of things.

So like the main one I had set up, it was all about like the leads that came into our system. So they come in, you know, through a web form or phone call, whatever the case is And I would create like my Google Drive folders and create like different lead documents, like pulling Zillow data that I needed or I wanted from like Elite Information. And it’s kind of like processing through different automations. I had Slack channels set up.

So like say we signed a project development agreement. Like I had a WINS channel and it’d be like, hey, we just signed a PDA for $10,000. You know, congrats. And kind of just those kind of things.

The second one I really built that was kind of popular was the I built this whole like dashboard system to ultimately connect to Google Sheets Preview. You know, before we really had dashboards and JobTread, but I had this Google Sheets. So this zap would like take in all the data from JobTread and essentially like build this dashboard real time so I could see a bunch of information around my leads, my sales, my like percentages, close rates, all this kind of stuff. So basically what you would do is like all the, you know, because I like, you know, we have clients that are like, oh, yeah, like, the first thing is I do.

I ask these questions. Then I go into to Google Maps or Zillow, and I look up their stuff and I. So basically, all those pieces that, like, most people listening to this are doing manually, you use Zapier to automatically do this So you weren’t having to click around and figure all the stuff out. It basically was coming to you almost like on a silver platter, where it’s like, this is all the information that you are about to manually do, but done automatically.

Yeah, well, I was using HubSpot at the time too, and I was like realizing, like, I’m entering here, I’m entering there. And it was like, this epiphany is like, why am I. Why am I doing this? I saw the Zapier thing.

I think it can do this kind of stuff. And it took a lot. It took a lot of hours. And that’s what I’d caution people is like, this stuff takes a lot of time and, you know, maybe you have like, the inclination to do it.

And I’m not like a techy person by nature. Like, I don’t know how to code, I don’t do that, all that stuff. But I’m like an inquisitive person and I hate doing things, like, repetitively. And again, I just like consistency of data.

And that’s really what I was focusing on is like, if I would touched it or one of my business partners touched it, like, it would still come out the same. And then last question is, what was like, the outcome? Like, was it like less stress over missing things? Was it like, this is just so much more efficient?

Like, what was like, the benefit of taking the time to set all this stuff up? Because it’s. I mean, you were maxing out at like, on Zapier’s amount of, like, how many, like, triggers could you basically have? Or how many things could happen?

Which is like I mean, we’ve used Zapier for Well, you know, up until recently, we, you know, I started using it eight, nine years ago. So I, I’ve, I’ve been in Zapier a lot. To have that many steps in a workflow is like, I’ve never even come close to that Maybe we had like 12 steps, but it would, it was like, yeah, even that was Would kill me. So you spent all this time to set this stuff up.

What’s like the ROI in the business for you? Well, again, for sure Right. But like, I had time invested. So like, there’s a trade off there.

But it’s. It’s like the what I was always looking for, especially the dashboards. Like, how can I get relevant information quickly to my fingertips without having to go dig and recreate and make another spreadsheet and spend. It’s like I could as quickly my spreadsheet and get information to get better feedback.

For example, like, I didn’t know at the time. Wolves are average zestimate of like a good prospect, you know, and ultimately came out as like, you know, just under like a million bucks of the house, like house value. And then so that became clear. It’s like, okay, say if I get five leads in a day, let’s say for example, and two of them are below that threshold, I would still call them, obviously, but like, okay, my prioritization is different now because three of the leads were like in the market of where I want to be.

So those get priorities. So it allowed me to like, make better decisions for which leads I want to spend like my better time with But the I. It’s like I didn’t like having to go to Zillow every single time and like, where’s the house? Let me see pictures, Let me.

How much is it worth? Things like that So I just wanted things like the lead sheet we created. You know, it’s like it would just populate. You know, I’d populate.

They would fill the form out. Great. I’d go to print it and I could call them immediately. Right.

You have to spend time, like retyping their name. Printing out other information is like, oh, Lee came in great print. Hey, hey, prospect, how you doing? You’re like And I had all the data at my fingertips to actually go act on Love it.

And Elliott over to you, like, give us your background. And like, how did you end up coming to this journey with Peter? Yeah, my family are craftsmen, generations going back. And I was the first one they sent to college.

So I did. I went for music education. That’s what I was into at the time. And I ended up being a band director five or six years teaching high school marching band.

Super fun. But going back to when I got my first computer when I was eight years old, I ended up being the family tech person, always troubleshooting the thing. When I was in high school, I taught myself how to code. I would play this game or like, get as many viruses as I could on my computer and time myself how quick I could clean the computer.

Like, I was always just Been a tech nerd leading into my band director job. I would build, I built like internal apps for the school. Thousand people, thousand high schoolers doing attendance and reward programs and some other grading type software internally for us to, for the teachers to use. And then when Covid hit kicked us out of schools, I went back to the family business, built internal tools for the construction company.

And we had an office manager at the time, said Elliott, you need to quit building these tools and let’s just find something that works. I said, okay, well it needs to have like an API. She didn’t know what that meant. And she looked around, found a few, did an interview with JobTread.

She loved it. She’s the one that signed up. And when we signed up for job trade, I said, okay, cool. First thing we need to build is a cash flow projections calendar.

And so I built, using job trades API, I was able to build this projections calendar visually to see like where all of our money is going in and out. And that’s really saved our butt over the last few years. And then from there just building tools internally for things that my, me and my company need and talking in the Facebook job trade, Facebook group, other people need the same thing. So just trying to figure out a way to offer that to them and share that with them.

Me and Peter met at the second JobTread connect 2023, 2024 and were talking about cost group templates and how to set up a budget. I had like this Google Drive of cost group templates that I would just email people who asked for it. And that turned into we built a website where people could just like sign up and get the folder of job trade cost group templates, which is pretty cool. It’s silly looking back.

And then me and Peter just nerded out about workflows and how to build our stuff. Peter used Zapier for all his stuff. I would just build it in code for like internal use. And then both of us just trying to be helpful and helping other people in the Facebook group.

And like a lot of people need this stuff, a lot of people want this stuff. So just figuring out a way how to share it. And that’s what turned into DataX. So it’s like a hodgepodge of tools that we’ve built for ourselves and we just try to share it with others.

And then how did you guys come up with the idea for this? Because like Peter was like, you know Elliott, you basically did everything by hand. Peter did it all with like calculators, like the I don’t know. The way that I would look at it is you want the hardcore, like, let me just figure it out without using the software.

Like, what then spurred this idea of like, hey, we could create this software company that basically combines both of Yalls brains and skill sets into an actual business. It was never really like a. The goal was never to be a business. I think we started it was like $20 a month.

We had a library of all of our templates that anybody could go download. And then we had these dashboards. Before job trade had dashboards, we needed to be able to see our data. I had my own dashboards and Peter had his own dashboards.

People were always looking for dashboards in JobTread and doing it themselves. And were like, well, we can do it and just let’s give people a place like log in, connect their job training seat and see their data. Visually. When we got like 10 people paying us like 20 bucks at.

First were like, okay, this might be something. And from there I don’t think there’s been a strategy other than to like help people and build tools that help other people. That is the strategy. Yeah, yeah.

Build helpful tools. How do you guys then like know, I mean like, you’re both in the industry, so are you basically just saying like, what are the things that we’re using in our own companies to then come back and go, hey, this is really helpful. How do we now make this thing production ready for all the users that are actually using it? Like, I guess how does your like ideation or strategy come about as far as like what you are building out and shipping?

I mean, it’s pretty organic, I’d suppose. Certainly the beginning. We’re solving our own problems, right? As contractors, like we suffered our own pain points and problems.

Like I built this whole zapier and like I hired this fiverr for like, I don’t know, 500 bucks to like give me basically like three lines of code. Which now is like, seems ridiculous because you know, you get a sneeze and chat GPT or Claude or whoever would just like give you this Perfect. So I, I was like telling, I was like, I paid this guy and this, that and he’s like, I mean you could like do it way more efficiently. Like, okay, cool.

How do I do? You know, let’s do that And so it became about like solving our own problems. And now it’s really like when we see things like users will tell us like, hey, can you build this other tool like for company Camp, for example, people are like, hey, I’d like to connect. Not so at the job level, but the customer level.

It’s like, okay, that’s pretty interesting. Like that sounds like good. And then a couple more people are asking about it like calendly. So it’s really probably a more organic, you know, we don’t have like some sophisticated roadmap.

AI takes the center stage on everything now. But I love like the toolbox stuff that we’re doing because it’s just so, it just works like, it’s just one of those things where people are like what they want out of certain AI tools of like save all my time, save do all this to make things better, more efficient. Like that’s the heart of like the tools that we build of like it just works, it works in the background. Like a perfect employee that never takes a sick day, never complains.

They work at 2am it doesn’t matter. You know, they just work. So it’s probably more organic than maybe it comes off. What would you guys say is like a big misconception that a lot of the companies that you that are using DataX like have around either what AI is or what it can do for their business?

Man. Well, Elliott can probably talk, I’ll save my first part and then Elliott can clean, can help bring it up. But I, what I’m seeing now is there’s massive fomo and basically people are like trying to solve problems that don’t exist in their business. And they’re trying like these, they’re trying to take huge bites, huge leaps like shoot for the moon kind of, it’s kind of like, hey, like, you know, but people are being told you’re going to be left behind, your business is going to close in six months because you’re not using AI.

So I think the FOMO is real and it’s like know just, it’s just calm down. Like a very small percentage of people are even paying for AI. So you’re still ahead of the curve. Yeah.

04% Of the entire global population pays $20 or more for AI. 04%. And then I think it drops. I think it jumps to like 16% of the total population has used like a free AI chatbot one at least one time.

And so like you’re drastically ahead of the entire world. So I don’t feel too far left behind. And I think another thing people kind of get wrong. And it’s partly because just how the consumer Models work.

When I say consumer or frontier models, I’m talking about like OpenAI or anthropic, like these big AI labs pushing billions of dollars into their products. And their strategy is to make the model smarter so the consumer us don’t have to work as hard to get the results that we need. And so they’re just pumping these data centers like full of resources to make the model smarter. So you can lazy type a prompt like, hey, write a bedtime story about Elliott and his dog.

Whereas, you know, these models have so much money and resources that it’s brute forcing that you’re a lazy prompt and figuring out what to do when. I’m a huge believer of like, give the models more context, like, who’s Elliott, what’s his dog, what type of dog? What do they do together? What kind of games do they play?

What does he like about the dog? Like give more context to your prompt so the models don’t have to work as hard. And that’s going to later like down the road. We’re kind of being like, me and Peter talk about like, the AI cost of AI could potentially triple when all these investors want their money back because they’re losing money hand over fist right now.

When these investors want their money back, easiest way to do it is to raise AI prices and so or cut your token uses, which is like how much AI usage you get it. And so if you can prompt better, it makes the AI results the same without spending as much tokens. So whenever the time comes the AI gets more expensive or whatever happens, if you know how to prompt and provide context really well, you’re keeping your AI cost down. You’re, it’s more efficient.

You can use cheaper models and not have to use the very expensive models where you can like lazy load, brute force the results. If you can give context and like really good prompts, you can use the cheaper models and still get the same results. So I think a lot of our support tickets at DataX can be resolved with like just better prompting. And that’s been a huge, a huge issue lately.

Well, and I want to get back to Peter’s idea of like, you know, solving problems that don’t exist. But Elliott, with what you’re describing on like the lazy prompting where it’s like, let’s say you just use claw and you go to, you know, go to JobTread and build out a new scope of work and there’s no context, there’s nothing. It doesn’t know you know, it knows what JobTread is, but it may not necessarily know where to go and find it and all that kind of stuff. Your whole example of like write a story about, you know, bedtime story about Elliott, like, is DataX not basically the same thing of what you’re describing?

Where it’s like, most people are going to go in and go, hey, use JobTread and build the scope of workout. But data X is like, hey, you just have to say, however, we have the guardrails where we’re saying, first off, this is what JobTread is Second off, if they say something like scope of work, this section of JobTread is probably where you could find it. So, like, is, am I interpreting DataX correctly? Where you guys are like, more of you have to say, like, we could, like, I know there’s a lot that goes into it.

Like, you connect a lot of things, but like, you’re basically helping out with the prompts to a certain extent to give the guardrails to people’s AI when it taps into JobTread. Yeah, and so that’s exactly right, Logan. Like our, the back end of DataX that no one can see is this massive system prompt. And we have several system prompts all put together depending on what you’re asking it.

But if you ask it to go look at your active jobs, we have a little prompt in the back end of DataX that tells the AI models exactly what active jobs means in JobTread. So it’s not having a Bruce brute force. Hey, okay, I’m looking at all the jobs. Oh, I see a custom field here.

I see closed here, see closed on date. I see open date. It doesn’t have to guess what fields to do. We tell it specifically inside of DataX.

Active jobs means jobs that haven’t been closed and they have an approved customer order. So it just knows immediately what to filter and go to it. And it’s, I mean, from our test, it’s anywhere from like 20 times to 50 times more efficient because we’ve predefined all of these tools and how to use JobTread and basic construction logic inside of the back end of DataX to help the models be way more efficient. And it’s really easy for us to do that on our side.

And it just helps all the consumers not hit their token limits be able to use, you know, cheaper models to get the same results. And like, from my side as a business owner, like, to me, I’m. Because we have everybody here and not that like, so Claude is one of these models and just it’s funny because at this point I assume everybody knows what it is, but like my brother’s a contractor and I asked him, you know, the other week, I was like, oh, do you guys use Claude? And he was like, who’s that?

And I was like, claude, like anthropic. He was like, what is that? I was like, it’s like chat gbt. He’s like, oh, I know what chat GBT is, but I’ve never used it.

I was like, so you don’t know what Claude is and you’ve Never actually used ChatGPT and he’s 30, he’s a like I assumed that like everybody knows what it is but I’m like kind of your point, Elliott. Like only a certain percentage of people like actually use it. So like I kind of realized that at that point. Like how does my father in law know what this is but my 30 year old brother who runs a business, a construction business, like doesn’t know.

So but it’s, you know, I digress however, I mean we use Claude and like I, we have everybody in our company has like the max plan. So like I’m less worried about the cost associated with it, but it’s more so the time efficiency of it of like are we getting it correct when we ask it to do these things? Can you talk a little bit more Elliott, about like the efficiency as far as like when you want it to do something, how data X can come in and basically make it more efficient but also like the first talk about like the frustration that happens with trying to get it to do something and it doesn’t do what you want it to do. I pushed an update last night actually and so we have a prompt assistant now so you can give it like your lazy prompt and our AI model runs over and makes the prompt better before sending it to the agent to like run the instructions.

The biggest thing is like giving it context and like giving it a set of instructions on what to do, what tools to call which, what items to look at in the job, tread in your job trade account. And giving it context could be as simple. Like you don’t have to go as deep as what I’m talking, you can give it context as simple as, hey, look at job 1213. I love the budget, I love how I built it out, I love the documents, I love the contract, I love how I did the schedule.

Use this job as an example to go create another job so you don’t have to give it all the Info you can just tell it where to look as like a first step of giving it good context. What was the first part of that Question was just about like how like So I think that answered the first part. But like talk about like the frustration that Okay, like I’m sure you can see like to a certain extent like what people are asking about and even with like the support tickets that you guys are getting about prompting like And I was telling Peter when Job Trade released, like the native, you know, tap into JobTread through Claude in my head I was like, I think at first there’s going to be a lot of people that go from, I’m paying DataX now I’m going to go back and do the free thing with JobTread. But I think ultimately this is going to be in my perspective really big for DataX because people are going to go, oh, I can now use this thing.

But Dan, this is so frustrating. For a hundred bucks a month I can remove all this frustration. Like a hundred bucks a month is nothing if it solves just a little bit that frustration. So can you talk a bit more about, I guess, the frustration that people will typically see when trying to use AI with JobTread versus DataX with JobTread.

Yeah, it goes back to us having these like 30 something tools defined with business logic of how to use JobTread, how construction companies work XYZ in the back end of DataX. So when you give prime, like if you’re just starting out with Cloud or ChatGPT and connecting JobTread and you’re saying like, hey, find my active jobs, like we mentioned earlier, it’s brute forcing and spending tens and thousands of tokens to figure out what’s working. And it might take 15, 20 minutes the first time you ask it. And, and if you ask it something slightly different, it can’t reuse that same memory and it’s got to brute force the next question and take tens of thousands of tokens and maybe 10, 15, 20 minutes to brute force its way till it gets answer and you’re going back and forth.

Maybe it gives you answer and it’s wrong and you have to like reset, like hey, use this filter or look at this type of document instead you’re pulling in the wrong to dos. Look at the to DOS assigned to Peter instead of Elliott, blah blah, and you’re like, keep going back and forth and then you’ve wasted four hours and you still don’t have what you’re looking for If you use like if you just go into DataX and ask it, you’re gonna have to give it a little bit more context because the DataX model is a little bit not, it’s not as smart as the frontier models. But if you get a little bit more context, you’ll get the same exact results because you don’t have to talk about the logic, just say, hey, Logan’s a new customer, he’s got this scope of work based on this email. Here, let me forward you, I’m going to forward the DataX agent an email and it’s going to read the email and get the context to go create the job.

Whereas if you’re trying to in Claude, it’s going to just brute force, it’s going to take probably three times as long in some cases. That being said, we understand that our model isn’t super, isn’t frontier level. So we just, we implemented this thing where you can use like your Claude API Token or your ChatGPT API tokens to use the frontier models inside the DataX wrapper or harness or whatever the new term slang you’re calling it. But the DataX software can use those other models if you’re not happy with like giving context and you’re okay brute forcing your prompts.

So there’s different ways to get about, get around it. We’re trying to offer solutions with the prompt assistant, with the bring your own API keys, with the template libraries, with all the agents that are pre built. You just need to go and edit some business data in the instructions. Like we hear all these frustrations.

A lot of people have forgotten how to like read and comprehend themselves like basic human abilities that we have trained ourselves to do for so long and we’ve kind of let some of that go and we just trust the AI models where it’s like if you just take a step back, don’t rush, provide some context, read actually what is giving you and give it back good prompting or good corrections in your prompts, you’re going to get better results. All the frustrations we’re seeing is like write a bedtime story about Elliott and his dog where it’s like there’s no context or build a new budget for a bathroom. Does it already exist in JobTread? Is there another client inside of JobTread that you have that’s already built a beautiful budget that we can just copy like there’s zero context given and that’s.

We gotta talk to like a human, like Train and talk to, you know, you can’t just tell a new employee, go build me a bathroom budget. Right? They’d be like, I. I’ll try my best, you know, and But it’s.

It’s like iterate with a human. So to To kind of like spit it back. Cause I want to make sure I’m understanding this and then make sure we’re on the same page. So basically what you’re saying here, Elliott, is like, DataX is really good at.

If you say, let’s say, you know, right now we have another budget or another example of a bathroom scope of work, I want you to go into JobTread, and I want you to replicate that for the Peter Ranney project that just came in DataX will say, I know exactly where to find it, what scope of work means, who the client is, all that kind of stuff. So DataX is really good at basically saying, if you give me instructions, I know the most direct path to get there. Because otherwise, and I experience this a lot, is like, if it didn’t have DataX and I gave that exact set of instructions, it would say, well, do you already have, like, okay, you said you have another scope of work in JobTread. I don’t know where to find it.

So let me keep spending time and over and over trying to find it. Hey, I found it. It’s not a bathroom, but it’s a kitchen. And it’s like, no, no, that’s not it.

I said bathroom. There is a bathroom in there. I promise it is So DataX has been good at that, but the problem that clients are getting is that they need to prompt it and go, hey, you already have it in there. DataX is good at getting it.

But for right now, people are saying, build me a bathroom scope of work without giving it any more. And what you just shipped out was like, all right, we’re going to actually make it. So you could hopefully make it a little bit easier and lazier to say, build me a bathroom scope of work. And it may ask you some questions, and then that way it gets you to the right answer.

And DataX can hit it directly. But it’s kind of like not a chicken or the egg situation, but it’s like, DataX is really good at getting you to the end path really quickly. You just got to get the right starting point, while the frontier models are really good at the starting point, but they’re just totally guessing what the pathway is and what the end result of the pathway actually is Yeah, that’s really well said. Okay, cool.

Thank you. I have. I have experienced a lot of the frustration around all this stuff too. Like, because we have.

We do. We have done a lot of stuff with AI at our company. And like, I was telling Peter, we have, like, this whole, like, client, and we said a podcast on it, like, this client brain where, like, every single one of our clients, we have, like, I could write now and go, hey, I want you to write an email coming from RanneyBlairWeidmann using their brand voice. And it’s an apology to a client that says, sorry, were delayed by one minute.

And it would. It has all the transcripts. It knows exactly what the brand voice is, how they say things, does Peter have a lisp or not? Like, stuff like that And it can do it that way, but I’ve done it the other way where it’s just totally guessing where they’re like, which Peter, is it?

Like, is it Ranney with an AI? It’s like a pain in the butt. So I. I’ve seen it both sides of, like, the pathway.

It’s kind of It’s funny because at first the frontier stuff is like, wow, this solves all my problems. This is great. But once you learn how to use it, the big frustration part is the step by step, like, how do we get from point A to point B, which is really where DataX comes in Which is why I truly think you guys are going to be in a really good spot. Because, like, once you have enough of that frustration, where it’s like, God, it keeps trying to get.

This is not even a Costco. This is just like, this is somebody’s phone number. That’s not what I’m looking like It just gets to that point where your pathway will become, like, the thing that everybody wants. I’ve got a.

Let’s add on to that real quick. Logan. I think I’m thinking in the future, Peter will tell you, like, it’s hard for me to live in the moment. I’m like, five years ahead all the time, and it’s really hard for me to live in the moment at all.

So I’m thinking and thinking ahead. We built DataX to be a platform that is agnostic to AI, to whichever AI model, because right now we’re using Claude and ChatGPT. You might be building projects inside of Claude and they’re awesome, but what if super duper AI.com comes out tomorrow and it’s way better than Claude? How are you going to get all your context out of Claude into super duperai.com you can’t.

It’s really hard. You can do it, but it’s really hard. DataX, you just flip a few triggers and you’re no longer using DataX AI or Cloud AI. You can a few buttons and you can get super duper AI.com to use your DataX software and all the agents and all your data that’s inside of DataX.

And so we’re really just building a platform that’s AI agnostic. So it can, I’m hoping I’m using that word right, so you can switch in and out the models. Because who knows, ChatGPT might be the winner this summer and cloud might step down a little bit and then super duper AI.com might come out in the fall and it beats everybody. And we just want to be able to be the platform that you can switch whichever model is the best at the time.

I think that’s something that people need to think about Like when you’re thinking planning long term, think about that So, Peter, something you brought up earlier was about the biggest misconception around AI was people are building or trying to solve problems that don’t even exist. And I’ve personally experienced this where I’ve. The amount of times that I’m like, let me just go to bed, it’s eight o’, clock, I get to wake up early and next thing you know it’s 11:30pm and then a week later, the thing I was working on until 11:30 at night was like this cool thing that has no impact on my life whatsoever. But Dan, this one dashboard that like, you know, told me like, what my like HRV on my whoop is like, it looks really good, but like, it doesn’t impact my life at all.

And it wasn’t even a problem I needed to solve, but I went down this rabbit hole of like, this thing is cool. Let me show my wife. And my wife was like, what? Like, that’s why you’re tired.

That’s so stupid. So, like, can you talk more about Because I see this a lot and I’ve fallen victim to it myself. Where people are trying, they’re using it to solve all these problems that aren’t even really problems. It’s like, your problem is not this Your problem is you don’t know what your numbers are at all.

Not what’s your speed to lead. You got no leads like that kind of stuff. So can you talk through what you meant by that? Well, like most technologies, they’re supposed to save us time and money and add more leisure to Our life.

And me and Elliott experience the exact opposite. You know, it’s like we are working like, you know, all the time. You know, one more prop and then I go to bed. But from a business perspective, you know, especially in the construction world and industry, like, you know, we solve businesses to run and that’s the most important thing is like run your business.

And where I’m seeing a lot of like the friction right now is people trying to blend their business practice with AI and their business practice may not be set up to solve the problems they think they have in their business. Like the biggest one I hear is I want just to estimate for me. Well, I mean we all get 20 emails a day from the estimator from, you know, wherever they are in the world or United States or whatever, you know, because that’s the pain point for us. But it’s like you gotta have your foundation of your company set up to allow for AI to actually like, if you want your, if you want an email every day of like, tell me if I’m on, off or on track on my schedules.

But your project managers don’t update their schedules. AI can’t help you, right? Like it can’t solve your problem. And back to the FOMO thing.

Well, first, like go run your business. You know, devote maybe a couple hours a week or something to this But, but more importantly, like there’s so like Elliott and I talk about this a lot. We banked her back of like, you know, different things. And it’s like you can buy bread at the grocery store, right?

Pre made, already ready to go make a peanut butter jelly sandwich. They also sell flour, milk, eggs and water. I don’t buy those ingredients, right? I go to the bread aisle and buy my bread.

So it’s kind of like this guy, you’re trying to build your own tools and that might be okay, but like there’s vendors and partners and other softwares that like that are already ready to go to help you. So like lean into those. And really, AI spend is not a cost anymore. It’s an roi.

You know, it’s, you’re creating revenue with this versus just like it’s a bottom line expense. But just don’t let the FOMO get to you. Start small. I was at a conference last or a couple weeks ago in Nashville and someone was chatting me in the hallway and you know about all the cool agents and things like that And he was, you know, this big grandson, says hey, just start with like this Agent, which I described to a grammar checker agent.

It just goes in and checks your grammar when you create a comment or create a day log. And he was like, no, no. I want like a bigger, A batter cooler agent. It’s kind of like, I get it, man.

But like, how do you climb Mount Everest, right? I don’t know how many base camps there are I’m assuming there’s more than, you know, somewhere around 10 or more. I have no idea. But it’s like, you can’t just get to the top.

Like, maybe that’s your goal to get there, but, like, start small. Like, take small pain points up. And I think the biggest opportunity to kind of round up my point here, I think the biggest opportunities, us as contractors with AI, is the opportunity to get reporting back to us, which we’ve really never had from a business perspective. Like, the structure of our companies are very like, do.

You got to go? Do, do. Whereas now it’s like with all my dashboards and zaps that I created, it’s like, now I can get information fed back to me to make better decisions to run my company. So don’t worry about the, you know, trying to create the next thing with Claude and all this Like, start with some reporting, some simple things.

Yeah, I found for, like, the stuff that we have built out, the things that we already have a process for, like a. Whether it’s an offline manual process or something that’s, like, not super efficient, but we actually have a process for, it is so much easier to build something out using some sort of, like, AI workflow rather than. Well, you know what would be cool is this thing that we have no system or process for, we have no data for But damn, it would look awesome if I could, like, build this thing out. And then ultimately, like, were getting stuck at, like, this 60 mark because we’re like, well, we don’t know where to go next because we don’t know what right looks like We don’t know.

You know, we also have no examples to show of what right looks like So there’s a lot of things that, like when. When I first got into it, because it’s very. It’s addicting. I mean, it was very much like, oh, we’re going to build this thing out.

And like, this is like, this is going to revolutionize the company. Like, CGN is just going to be Logan and 45 agents. Like, you know, just like, silly stuff. But like, that’s how, like, at first, it’s like, I’ve never felt so empowered that like it’s a visionary world now.

Like, I don’t need something to do all this stuff. I just need Claude always on and like ultimately like when it’s actually like, oh, this is like a, like in the, like all the sales calls now like automatically get, they get transcribed, they then get put through a full, like we have a, like a 13 page doc of how an initial sales call should look and it compares it based off of that and then it shoots the answers into Slack and says, this is how it is This is what you did really well, this is what you did differently. And then it has a button that says, do you want me to draft the follow up email from this sales call? So it’s like that was already something though, that like I wasn’t saying, hey, make me a sales process.

It was like I had to spend time to go back and forth and go, all right, first off, here’s all these sales calls that we have. These are the ones that worked really well and these are the ones that didn’t go well. What are the differences? Great, let’s start there.

And then it’s a lot of back and forth to figure it out. But it was because like we already had a process in a system around how the sales used to go. It was so much easier to build this thing. And the whole thing was built out in basically a day.

It was a long day. Cause you had to first document the whole sales process. But once that was done, it was easy. I think reading comprehension is the skill that people have to bring back.

Right? It’s, it’s hard because contractors in general, like we are doers, fixers with our hands and we’ve, you know, reading comprehension, probably for most of us it was not like a strong subject in school. And now like you described, you had to spend a majority of the day reading and comprehending and going back to like, no, I want this, let me read this Oh, that’s a good idea. But I don’t like that idea.

I actually want this other idea. So that’s, it’s a hard skill because it takes focus and AI is making us not dumber, it’s making us lazier. So we don’t want to actually read and like converse with it. So something.

Because I agree with all that because like the critical thinking is a big part of it. Because like the nuance, because it’ll give me something wrong. And I’m like, why did you give me wrong? And it’s like, well, the first sentence it said speak professionally.

And then in sentence 45 it said make sure that you use Logan lingo, which is more impersonal. So there’s elements there that like you kind of just say, create this prompt. Now go do it. This thing sucks.

Why does it suck? But with like that specific workflow that I went through of like sales call into, transcribed into, you know, it goes through all that stuff. That’s to me, one of the biggest benefits is like it connects all these things together, but it’s also smartly connecting all these things together. So one of the things that data x, in my opinion, it’s like one of the strongest things is the fact that like somebody who has a phone system doesn’t then have to go, well, I have all these phone transcriptions.

Do I now upload this into this? And I do this into this Can you talk about like yalls approach to like connecting all these apps together to make sure that like it all syncs up and it’s like more harmonious rather than you have all these different softwares and these different texts that you basically have to manually try to piece it all together. There’s two parts to that Elliott will dive in deep. There’s two parts to that, like the connecting database and like this new UI concept.

You want to talk? We. We love talking about this Yeah. First, you gotta make sure that you have a central hub where all of your stuff is Logan, you mentioned a brain.

I’m sure it’s like the second brain idea. Yes. Going around, you know, this is It’s just a database of all your knowledge. I think contractors, most of us, could use JobTread as your central hub database of all your knowledge.

Right. Just figure out how to get all the data in there. What you’re talking about like connecting databases together. Right.

Like this Got your Gmail, your debt, your JobTread, your Google sheets, your all, you know, your phone systems, your stuff, all these things are now connected and your AI can have like really good context of like where these different things lie. And I don’t know if it’s, you know, too futuristic or what, but like we got some good plans with the power of the MCP to really open it up to where any software with MCP can be connected to your, in this case, your source of truth being called JobTread. Whatever your source of truth is, Google sheet or something, if it has mcp, they can talk. Right?

And that’s where the power is because no longer these like everything’s on siloed and it’s on an Island. You have to, like, copy here and paste over there and And you kind of do this whole thing back and forth. But I think the second point is now your databases are connected. Like the UI of how we experience software.

We got to go click buttons and click here to get there, click there to get that As a human interaction, AI can do those things for us. So it’s like we’re just interfacing with a chat and saying, I have this budget or I have this scope of work. I want you to create the job. I want you to create the budget.

Look at, right? Giving all the context in a chat. And then you go make a sandwich for lunch and you come back and your JobTread’s kind of like built out. That idea of connecting data, though, that, like, that happened before AI.

Like you mentioned Logan, about, like, phone systems. Like, we, like, me and Peter both use Quo, Dominic uses Dialpad, and other clients of ours use RingCentral. So, like, the idea is like, after my phone call on Quo Cuo already does the summary and the to DOS and the after meeting to DOS and all that stuff and the transcript and the audio recording, I was like, okay, if we could push this into JobTread, then my office manager and project manager can see the conversation I just had as the sales guy. They can see the same conversation in JobTread and not have to go to a different software.

And then talking about the Zillow connection. If I create a new job and we pull in all the Zillow data, everybody can automatically see that in the job field, people are project manager, sales, office. Everybody can see that in one place. So the idea, like, trying to get everybody to use a central hub.

Try not to have your project managers and office managers have to use different software to see the same data that could easily be inside of your JobTread. So it started as, like, you find a central hub. For us, it was JobTread. Find a way to automate all your data in there.

Now that we have AI, just have AI look at all the data inside your JobTread and help you make decisions. The. The. I mean, I went to an experience one time where were about to Before we, like, remodeled or we added a bathroom, were on the phone with We’d already talked to the salesperson.

We went into a design agreement, and we talked to the designer. And the first call, the designer, you know, we talked to the salesperson probably over the span of, I don’t know, two hours and all about how we’re adding a bathroom to a room. And we get on, and they come in and they. They scan the whole house and all that kind of stuff.

And we get on with the designer and And the designer pulls up an existing bathroom and goes, is this the bathroom that we’re remodeling? And we’re like, what? No. And she goes, oh, sorry.

Is it this one? And we’re like, we’re not remodeling a bathroom. We’re adding a bathroom. And she goes, I’m sorry.

Like, I’ll let you guys go. You just talk about your project. And that to me, we end up not using that company. The owner was great, called me.

So I’m so, like, what happened? Like, let’s go through it. Was. It was It all resolved itself.

But basically my wife and I were like, man, if this Salesperson didn’t spend 10 seconds ago before we hop on the phone, they’re not remodeling. They’re adding. That would have gotten us 80% of the way there. Because they would have said, where are we adding this bathroom?

At least we know that versus, like, which bathroom we’re remodeling. And that, to me, was like, such, like, that was the first time. I was like, you know, be really good if they just had some sort of, like, recording system where the salesperson gets the notes over to the designer gets the notes over to the PM or whoever’s running production in one seamless path. Like, that would solve so many problems.

And if everybody has it now in a centralized location, God, like, that would be huge. And that was like, as a homeowner, that That. That was a big part of what lost the deal was the fact that it was some other things with the salesperson and stuff like that But ultimately, that was like, kind of the final straw where my wife and I were like, I just don’t. I don’t feel the love on this one.

Like, it just seems like maybe this is too small for them and that’s why they didn’t bring it together. So to me, when we’re connecting this data to then put it into JobTread and to have it all centralized, like, to have that one, like what you guys call, like, the one point of truth to it all. I mean, that makes it so much easier for everybody to have the full flow. Where.

Because we experience this as a company ourselves, where it’s like we have a. Because all of our calls are transcribed and gets put in the client brain and stuff like that Where if a client one time just goes, hey, by the way, like, I like the color green. And I hear it on sales call a month and a half later when they’re doing the actual website and we had one speck of green, they’re like, I told Logan I didn’t like green. And it’s like in the span of four hours of conversation.

I’m sorry, I forgot to say that But with our client brain, it would have triggered to the project manager on the website side. By the way, when you’re building the website, they made a comment. It was a short comment. They do not like green.

Confirm with the client or just don’t do it. That to me is like a huge power of all this stuff that like we’ve experienced on our side, that solves a lot of these little communication problems that I know in the construction space is like super prevalent as well. Well, you’re describing, in my opinion, like the most, second most important use case for AI is like improving the customer experience. Right.

We’re all looking as, how can this solve my problems? How can it fix my issues? And that’s, you know, not to downplay those, but it’s like all this should be for the benefit of the client. Like, have you heard the Jeff Bezos thing where they always had an empty seat?

I don’t know if it’s actually true or not, but they had an empty seat at the table of their meetings. The empty chair represented the customer. Right. So it’s like they’re making these decisions, talking about things, but it’s like, how does this affect the customer?

You know, so that’s especially in the sales perspective. And then you get to the project and you know, the project manager shows up day one, and you know, hey, Logan, which bathroom going? Yeah. You know, and you’re like, we’re adding a bathroom.

And it’s like, oh, okay, let me read my notes here. And yeah, so that’s like improving the customer experience is like another always should be like at the front of what we’re thinking about how to improve our company. What would you say is like the The most used workflow or framework or use case that your clients use within DataX? I think it’s the Peter, correct me if I’m wrong.

I think it’s the receipt processor. Yeah. Right. So like, that’s been an issue for like a lot of people.

Logging receipts into JobTread is Is always a huge task that you have to read the receipt, you have to figure out the cost items from the receipt that match in the The job budget you have to cost it all correctly. Create the vendor bill X, Y, Z. We have this data X agent that is called receipt processor. It has its own email address.

So anytime you get a receipt from Floor and Decor, Home Depot, Ferguson in your email, you can just forward the email to the agent’s email and it triggers it to do all the work backing up a second. Our data X agents work based on triggers. Whereas like Most people use ChatGPT and Claude, you like type, then it does something and it doesn’t do anything again until you type. Logan, you mentioned you wish you had a quad on 24.

7 And that’s kind of the idea. We have the data X agents, they’re on 24. 7 Trigger working when it like while you’re sleeping. And so it has an email address.

You forward the email to this agent and it grabs the receipt off the attachment of the email. It reads the email. If you added context in the email, it creates the vendor bill and job trade. It finds the job, it finds the cost items in the budget, matches those with the cost items on the receipt, creates the receipt, attaches the receipt to the vendor bill, marks it paid, boom, you’re done.

That happens in the background while you’re working on other more important things. I like to do it at the end of the day. I send 20 emails at the end of the day, all my receipts to it and then I log off and I go make dinner and play with the family. And when I come back right before bed, I’ll like check it.

Okay, cool. Everything’s job costs are great. I’ll see you tomorrow. I think the receipt processor is the biggest one.

Probably just because I talk about it so much, everybody uses it. And then the second biggest one I think is starting to be our grammar checker, the super easy agent that Peter mentioned earlier. Anytime a daily log or a comment is made that’s client facing the AI checks it for grammar. And a good example I gave, I told Peter about the other day was we had a client actually, sorry, we had a project manager asking the client, hey, what color was the bathroom?

And his comment was bathroom, color, question mark. And that’s all he asked. And sure you could gain, you could figure out what he’s asking about But again, help the human brain not have to work as hard. The AI went through and said, Mr.

Client, what color do you want? Bathroom 3. Is it this color that it pulled out of the budget and it with a question mark? And the client was like, oh yeah, that’s the exact color I was looking for, rather than the project manager’s bathroom color question mark.

But the AI knew, based on the schedule, based on the budget, it knew how to ask the question better and gave more context to the client to get a better answer. And no one had to think very hard. Yeah, I think that’s. That’s starting to become number two.

Well, I think that I mean, just that example right there is like, to me, as a homeowner, that solves two problems. One, it now makes the project manager seem more white collar, if you want to call that, or educated, which makes me trust them more. Like, if we did a podcast yesterday with one of our clients, and they will, you know, come out at the time of this recording probably last week or two weeks ago. But basically, he works in I mean, his clients are like, probably the The.

My guess is, like, the The poorest client that he has is probably worth 20 million bucks. Like, they were He were And a lot of our billionaires. So he works with the elite of the elite, and they’re not wanting to hire this chuck in a truck that’s just ripping cigs that’s like, hey, what’s up? Like that It’s just not like that So when you have this grammar checker that’s now like, hey, Logan and Audrey, you know, first, there’s deference associated with that I feel better that they’re, like, on my same level when it comes to, like, education, because in society, we still put a big love, you know, emphasis on, like, the education level to see, like, what’s your status and stuff like that So there’s all those things that go into it, but then you flip it around into, like, instead of bathroom color.

Like, we talk about that sometimes at our company here, where we’re like, if I went to To Peter and at this point and, you know, we’ve been with RanneyBlairWeidmann for three or four years, and I was like a new. You know, if I was the new account manager on Peter’s account, I’m like, so, Peter, what kind of work do y’ all do? Like, his first thing is not going to be like, oh, what an inquisitive young guy. His first thing is gonna be like, does this guy even know anything about Click.

Yeah. So it’s. It’s this level of competence where it’s not like, bathroom color. And I’m first thinking, like, first off, do you not know that we’re doing a bathroom, let alone, like, we talked about this?

I told you, I didn’t like the color green. So if it’s now going in and pulling it out going, hey, maybe this is the wrong answer, but at least it’s showing you that Like I referenced something earlier and said, is it this bathroom? It makes it easier. I’m way more likely to respond.

I. Otherwise, if I get bathroom color and if I go on this rabbit hole of like, first off, they know we’re doing a bathroom. Second off, they know that we’re painting. Third off, let me just call the owner.

I don’t think the owner is going to love getting that call to go. They look, he’s just not good at communication. Sorry about that You know, so like, it’s such a small thing, but as somebody who now has done several large projects at my house over the past few years, that stuff, it’s small, but it goes such a big way when it comes to like trust that I just assume that they have it covered versus like, okay, what’s going on here that you would even ask me? Bathroom, color.

So I digress. Well, I’m glad you see the, you can see the vets, even that it’s a deal. I mean, I talk to people on the phone and I’m like, I see how you email. I gotta be blunt.

My wife would not, she would not like that Like, you’re not. I get it, but my wife doesn’t get it. Like, she is a banker. She’s your ideal client.

She’s the one with the money. She’s the one that like has the taste, all that kind of stuff. And if you’re like, you know, my dad used to do this He was a contractor and like, he would get a inquiry from like a client and it’s like this long. Like, I’m so happy with this It’s all this stuff.

I just have a few questions. What about this? What about this? What about this?

Sincerely looking forward to your conversation. So and so And my dad would respond with call me, period. And I’m like, dad, at least fake it, man. Like, at least respond to a couple questions like, or just call them.

But like, it was just not an appropriate response. So something as small as that in the eyes of like a discerning customer, maybe if it’s a. Look, I hired you just to come out and you know, put polymeric sand in my patio, maybe that’s fine. But if you’re doing like higher end interior remodeling, home building, anything where it’s like you want to trust the person doing it, you gotta have that stuff put in place.

That’s a big thing. So let’s do this I would love to hear like, Peter, what is your favorite feature? I would say, like, because I think most of the battle with all this AI stuff kind of to your point earlier, Peters, people don’t know how to use it. Like, they know what it could do, but they don’t even know what the use cases are So do you guys have like some other cool, like, use cases that you’re like, we got the grammar one, you know, we have the receipt one.

What else are some cool things that you guys are Are seeing people use it for? Because I think once you can start to see, you know, what the other side of it is, it makes it so much easier to go, wow, that’s a great idea. I never thought about that, but I can’t wait to implement it. Peter, you should talk about the one we did this morning.

No, I’m blanking with the change order. In the cash flow account. Oh, yeah. So we had.

This is a, this a combination, you know, so of tools that data offers. So the cash flow calendar we kind of mentioned earlier and this particular customer data customer, she was looking to have like she sends the invoice and to And then ultimately for the cash flow calendar work, you put a task on the schedule and JobTread. And that task then populates into the data X calendar so you can see your cash flow on a weekly or monthly basis kind of deal. And so she was, we helped her create an agent.

So you would manually send the invoice typically. Right. Which is probably a good thing. You click send.

Now this agent runs in the background. We set hers just up for change orders. This could be extrapolated out to any customer invoice gets sent. Then the agent runs and says, hey, this invoice was set.

Go ahead and add the task on the schedule. And so it does that automatically. And then if for some reason maybe, hey, actually we’re delayed on the project, or let’s do it the other way. We’re ahead of schedule, right?

We’re gonna. We need to collect that money earlier. You drag it, you know, a week before, and then that is reflected back in the cash flow calendar because the agent is seeing that change happen and then it’ll go through and update, which is pretty cool. A simpler one, I would say I helped someone set this up a couple weeks ago.

They were, they had a web form on their website that was not connected to like any, you know, just one of those like they just get an email type thing, right? So probably pretty common among our industry and that’s totally cool. This agent, because our agents can send and receive emails. This agent would get the email of the contact and create the customer and the job if you want it to But their case, they just need the customer created.

You’ll go and create the customer, their contact record, get their email, get their phone number, drop that in and whatever project description was in the this email. Bam. So this is like a simple thing, right? So like not saying get rid of your admins now your admin can do something like client facing, right?

Instead of just keying words into JobTread right now they can do something a little bit different. Our like outside of the ones Elliott mentioned, I think the probably the biggest one still out there off on the board. People want us to figure out how to help them build their estimates. Currently there’s not a great way for AI to like read plans because most plans in residential aren’t developed enough to like what do these lines on this thing mean kind of deal.

But we’re seeing like one use case I saw it was a deck builder and he has a Google sheet. He goes to the prospect’s house, said, you know, you want this size deck, this material, this kind of railing, this kind of roof, blah. So he spreadsheet like on site, he emails it to this agent essentially as he, you know, call it driving back to the office or whatever. This email this agent is working, it’s building his budget, building his schedule, developing kind of a list of questions that may be a concern.

Kind of looking back on some past jobs. Do you have enough screws? You have enough, you know, do you have enough time for your porta Potty, how many dumpsters do you think you need? All that kind of stuff.

And so the agent was like building this estimate. What I, I think the For most contractors like that aren’t single trade like deck or something like kind of like more, a little more straightforward. They can use it an estimate builder help build their like prelim budgets, like their ballpark estimates, you know, kind of quick, you know, cursory, you know, cuts at the scope of work. Scope of work is a good one too.

Can totally do scope of work. But yeah, and then with all this like the one thing is I, I think people look at it as like, should I get DataX or should I get Claude? DataX is like, it’s like to me it’s a tool that works in a very specific way. But like Claude, I mean like I’m running a race.

I have my own dashboard that connects my whoop with my trading schedule with, you know, the calendar with all like all this stuff. So it’s like nothing in DataX would be used for that So I use Claude for that But when it comes to like the business, like you guys have so many use cases on your website that says this is what this does, this is how you do it. Like, yeah, to me, running, it’s a ultra. It’s a backyard ultra.

I know I am. You’re one of those. I have with it with two kids, I’ve been slacking on going to the gym. So I needed something to like sign up for and I want to sign up for a 5k.

So I have a backyard. So I slept on my 2 year old’s floor last night and then Woke up at 5 this morning and ran. So. But I needed to because I don’t want to show up to this race and suck.

But you guys have all these use cases that it’s like, you know, and I see it in like the JobTread Pros community where it’s like, well, hell, like I was talking to Eric about this, the CEO of JobTread, but like we had a client that we. Built a website for He came back and said, I need more leads, I need help with the marketing. And we talked through like their full process and their full pipeline. And the issue is not that they need more leads.

The issue is it was taking like 11 days to build out like a preliminary scope of work for clients. And I was like, my goal is to get you busier, like not slow this thing down. You need to get job trade dialed in Have you heard of Claude and DataX? And they were like, no, what’s that?

And I was like, you don’t want to sit here and learn how to do it all yourself. You just want to be able to go in there and you even have it now where like we have at our company we use this thing called Whisper Flow, but basically it’s like the voice dictation that I just speak into it. My wife thinks it’s weird because we’re watching TV and I’m like very like trying to like quietly speaking to my computer. But it’s like a sensual voice.

So she thinks I’m like romanticized with Claude. But basically you speak into it and it makes it so much easier where it’s like, hey, I need this and this Can you do it? It pops it in and it goes. So I was like, to him, I was like, I don’t think you need me at all.

I think what you need is you need to get this thing dialed in And instead of you having to basically learn long division on your own, there’s this company that they’re. They’re the The calculator. Like, they already figured out how to do the math. You just got to push the buttons in and it’s done for you.

So I’ve been trying to get as many of our clients onto DataX as I can, because we need it for cgn. We need people to have accurate scopes of work because we tap our client brain into JobTread to then pull out scopes of work, cost catalogs, if they have transcripts, all that kind of stuff to feed our own client brain. So when I’m writing a blog for Peter, I can look. And I’m not writing a blog that’s like, what does a $30,000 bathroom remodel cost?

Because, like, that’s. They can’t do it at that price point. But it would be like, what’s the difference between $125,000 and a $200,000 bathroom remodel? Because that’s more in line with, like, his budgets.

But we know that based off of their data. But we don’t know that if they don’t build all this stuff out. So I get very fired up about it because I see the use cases for it directly helps us. So my My 2 cents on all this is like, you don’t want just like one or the other.

It’s not an either or It is like, you kind of need both, and they have two different purposes to it. And your way is just so much more efficient to like, actually be able to use your data in a and Which I think, I guess wise you guys call it a DataX. Not just like, how to create a scope of work X, but it’s like super intuitive and cool of what you guys are doing. Yeah.

And going back what I said earlier, like, if Claude gets dethroned as like, the best AI and super duper AI.com comes out this summer, and it’s awesome. You can just use that Plug that into DataX and your workflow doesn’t change. Right. Just the engine changes.

Yeah. Here’s what I’d love to do is if you guys, if people want to get a hold of DataX, we’ll put it in the show notes. Is it DataX to Okay, give us your final parting words. When it comes to what do you see over the next year as far as what’s going to be really important to the construction industry when it comes to knowing or using AI?

I think I’ll go first, Peter. I think a lot of people are trying to use and we kind of touched on it. A lot of people are trying to use AI to do the stuff and I think it’s, it does it really well and it’ll brute force it if you’re using a frontier model. But maybe we all need, especially in construction, we need to shift our brains a little bit because construction, it’s very client facing.

Humans still have to do stuff in the human world. I think a better use of AI would to be provide all the context that you can to AI and have it tell the humans what to do. So if like every morning you got an alert say hey, office managers need to do XYZ. Project managers need to check on jobs 1, 2, 3.

For these three issues, admin need to do this Field crews need to do this If you provide all the context to the AI and have it tell the humans what to do, I think that’s going to be a better use case long term rather than trying to brute force. Hey, do this thing in the company, hey do this client facing thing in the company where there’s like a higher chance of error because it is just AI. It’s not, you know, it’s prone to mistakes. But if you can give it context and tell the humans what to do in the real world, it’s super smart and it knows what to do.

But if it can tell you what to do and the humans go out and do those activities, I think it’s a better use case long term. Love it. Peter went in, but I can always add something. Yeah, I guess top that All right, I got you on that Yeah.

You know, again, I’ve said probably a couple times like, you know, start small, like don’t be. It’s, you know, I’ve been to a couple conferences about AI and actually we left this guy this really cool presentation and demonstration of his AI. And then were talking later in the hallway as were kind of leaving the conference and he’s like showing me his phone and like all the different things that he has to go read and answer because his AI has been like asking him all these questions. And so it’s kind of like you see the one side but like everything has a double, you know, there’s two sides of it.

So focus on running your business. Don’t worry about like everybody else. I would say, like Like any problem you’re trying to solve, like, hone in, like, the, you know, the five whys. Like, get to the root problem of your, like, the issue.

Maybe AI can solve it, maybe it can’t. But AI can’t solve all your problems yet. And you still will be in business if you know, a year OR 2 from AI is okay. Tractors.

19, 20. Right. 20. 30% Of us were in agriculture, and we’re farmers going, this tractor’s gonna, like, decimate the whole thing.

Less than, like, 2% of the population’s in agriculture now. Right. It creates new jobs. So it’s like the new opportunity, the new frontier.

Like, be open and optimistic about what AI can do for you and your business. Yeah, maybe. Maybe. A lot of people end up being construction for data centers.

You never know. This will be on the moon. I don’t. I don’t know if I want to Yeah, I know.

And be a contractor. All right. Peter Elliott, gentlemen. Thank you guys very much.

It’s been a pleasure. Thanks. Thank you.