Episode 68: AI, Hiring & The Future of Lean Land Investing

Clayton Hepler (00:00)
welcome to another episode of the ground game podcast. This is your co host, Clay Hepler.

Justin Piche (00:08)
And this is your other ghost, Justin Piche and we're here to show you how to win the ground

Good morning.

Clayton Hepler (00:27)
Dude, mad respect for you this morning. You are up early. I'm up early, but you are up early. So are you normally an early riser on the weekends? You know who's texting me right now? Literally Ben.

Justin Piche (00:35)
⁓ man, last...

Who?

Baker?

Clayton Hepler (00:48)
Yeah, he's literally texting me right now. I'm talking to him. I'm literally talking to him.

Justin Piche (00:54)
Yeah, let's see, he's probably texting me too. Nope.

Clayton Hepler (01:00)
Long sir no no in a good way so long so he's

Justin Piche (01:04)
So Ben is texting you at six in the morning for him.

Clayton Hepler (01:08)
Yep.

Justin Piche (01:10)
haha

Clayton Hepler (01:12)
⁓ the dude, I love that dude. ⁓ I know you, you and him are partners. He's just a good guy. He's just a good guy.

Justin Piche (01:17)
Yeah, he's a good day. He's awesome. ⁓

man. I am not normally a morning person on the weekends. Like I typically will sleep until like I have to get up because one of the kids is up essentially. But, and I'm also a person that likes to go to bed late. It's interesting. I like to go to bed late because I like to do things in the evening that I like want to do and

Now that I have three kids and one who's going to be born in the next two weeks, almost four kids, I don't have a ton of time anymore. So when I get up, yeah, shocking, shocking, shocking. Get up, help get the kids ready. I'm really not at my computer until nine.

Clayton Hepler (01:58)
Shocking, shocking.

Justin Piche (02:06)
you i'm actually obviously get the on it earlier if i need to but it was good meetings before nine so i'm not like really on the computer of the reviewing some emails are but looking at my calendar and knowing what i'm gonna do for the day and and i'm i'm basically nonstop from nine to five and i don't eat lunch and just like i'm i'm just on calls like whatever and then but at five like that's the limit

That's like, if I'm not out and like hanging with the family at five, then I'm, yeah, I need to be doing that, right? So, I mean, sometimes it's a little bit earlier. Sometimes it's 4.30, but typically it's five or just like a little bit after five and then I'm off the computer, we're out, you know, I'm playing with the kids, cooking dinner with my wife or she's cooking dinner, I'm playing with the kids or we do after school stuff with the kids. And then ⁓ bedtime.

Clayton Hepler (02:43)
Yeah

Justin Piche (03:03)
I'm usually down stairs by eight at night after talking to the kids, maybe 8.15. The kids are totally not coming out their rooms and asleep. And then it's like, okay, well, I worked and I didn't really hang out with my wife at all. I kind of hung out with my kids. So that's the only time to either hang out or do anything else I like to do, like play music or mess with my pedal board stuff or.

do the things during the day that I didn't have time to get done but have to get done or I feel like they need to get done. Yeah, not yet. Not yet. It's bad. But there's some good news on that front. I just built my deadlift platform, squat rack platform in the garage. I used to have one for a long time. I worked out so consistently, it was ridiculous. And then when we moved out of our house before construction,

Clayton Hepler (03:35)
When do you work out,

Justin Piche (03:58)
they demoed the squat rack platform and I moved into a rental house and it was like we were using the garage for storage essentially because we had more stuff from our regular house before I moved into that rental so I had no garage gym and we moved back here in July and it should have been the very first thing I did but the garage was just full of stuff.

And so finally I got like the garage storage solution. I've got like the racks that come up on the ceiling and like the racks, yeah, dad mode engaged. Like we got rid of a bunch of stuff. And so I bought all the wood at Home Depot. I borrowed a buddy of mine's truck, went and picked it up, put it together the other night. And I was literally about to assemble the squat rack. And I realized I'm missing one of the gusset plates that holds the vertical section of the squat rack to the horizontal section.

Clayton Hepler (04:24)
Dad, dad mode engaged. Yeah, dad mode engaged.

Justin Piche (04:48)
Like I've lost it in the move, there's four of them, two plates that go on either side. And so then I called my welder the next morning and sent him a picture of it. And I was like, hey, I need you to make me another one of these. Cause like, I can't buy it. There's no, this company that I have this rack from like doesn't exist anymore, but it's a really nice rack. So I need you to make it. So he picked it up two days ago. He's supposed to come back tomorrow to like paint the rest of my fence. And I'm hoping that he'll bring back my gusset plate with a new one that he made because it's just a piece of.

pretty thick, I don't know, this thick sheet metal with four holes in it to hold the whole frame together. Once he does that, then I can read some of my garage camp. So anyway, that's a really long story to say. I should be working out soon, although I have a baby coming and then I'll have no sleep.

Clayton Hepler (05:32)
Yeah dude, my gosh.

You got some stuff going on on the horizon here.

Justin Piche (05:36)
What's going on with you?

Clayton Hepler (05:39)
we are going actually after some mobile home on land packages.

Justin Piche (05:43)
Okay, dude, which I think is, by the way, I think that's a great idea this time. in this time in like the industry, think it's like offering value add is a great option.

Clayton Hepler (05:58)
Yes, so we're doing that, which is, I have, it is a testament to hiring really good people, Justin, that I have a team member in my marketing department who is a superstar. She is a superstar. And it's hard to find people like true A players.

there there there's always conflicting schools of thought which is you build teams on when we when I say B players, it's not a pejorative thing. It's like, you know, people that are not like, high high engine people, right? She has a level she's an engine that's like, above my head should like like, she's the hardest working she's incredibly smart.

And she's just eating this entire initiative up and we're going to learn about it, which is really cool. So we're actually putting, we're targeting properties now to go after it, which I'm really excited about. We've gotten a ton of offers lately on, on properties. and so that's really, that's really cool. ⁓ I'm in search of an acquisition manager and,

I can bring this to this call. think it could be really interesting to actually talk through in real time with you. Because it's, I'll save that for later, but I think it's really relevant to the topics you and I actually texted about earlier. And so, but that's kind of the big thing for me looking for that AM. And then I'm doing a big webinar.

next week, so I'm preparing for that for the state of the industry, right? Like, so we took all of our data from our business in January and December and all of the data from the people in the land, the deal engine community to figure out like what it what actually is working.

Justin Piche (08:04)
Hmm.

Clayton Hepler (08:05)
Like what's working now? I know you guys, you did a podcast about that. On the REtipster maybe it was REtipster. You did a podcast about that? Yeah, yeah, yeah.

Justin Piche (08:15)
It was like the round table.

Yeah, I was part of that round table. Yeah.

Clayton Hepler (08:20)
And so that's,

that's it. And then I'm launching actually a challenge on social media, 100 days to 500k in profit.

Justin Piche (08:28)
Realize or okay. Yeah, yeah from now. So you got a double close essentially

Clayton Hepler (08:30)
Realize. Realize.

from,

well, it's deals that I've had under contract as well. What I'm trying to show is that it's still possible to make money.

Justin Piche (08:41)
Okay.

And it is, it just, yeah, it definitely is.

Clayton Hepler (08:53)
How about you, man? What's new with you?

Justin Piche (08:56)
let's see, on the business side...

We're starting to sell lots in one of our funds, so talking about Ben Baker. Maybe a step back. We formally formed our investment firm partnership called Triune Capital Partners. ⁓ That is formed, which is great. We have our website. ⁓

kind of starting to consolidate our existing funds under Triune Capital Partners, like our existing management company, we're moving under the management company of that entity and all kinds of, we're just getting everything kind of situated correctly for like a formal partnership rather than just like me and my partners both owning ourselves shares of like a GP and so kind of more for our structure, which is great. We're starting to sell lots on

our Spokane development, which is great. We project out, hey, we get this property, we get these properties purchased under contract, and we don't start to sell anything for six, seven, eight months, typically, is what we plan. Because there's time, we gotta market this thing, we often have to do work to it, it takes time to get those things done and approved. And on the first deal that we did, Burn It, we're behind.

Clayton Hepler (09:57)
nice.

Justin Piche (10:24)
that metric. the reason is we had a huge regulation delay with ⁓ the Lower Colorado River Authority, LCRA. They're a Texas government board that prevents, that's goal is essentially to prevent waste runoff into water systems like rivers and aquifers and whatever. And so they monitor construction debris, they monitor

⁓ They're responsible for, they're the ones who sign off on all the silt fencing design for major road and construction projects, They're the ones that make sure any sensitive slope areas that drain into some sort of drainage basin are protected, that nobody can develop or disturb those slope areas. And it just so happens that every employee turned over on their team. Nobody on the team had any experience doing anything in this county.

Clayton Hepler (11:15)
⁓ my gosh.

Justin Piche (11:20)
And so when our engineers submitted our plans to them, the normal way he's done it for the last 20 projects, he's gotten approved for them. They had all this pushback. And so we had to convince them that our way of doing it was the correct way. So what they wanted us to do, and I'll get back to Spokane because I just went on this crazy ride, but I'm going to take you guys on a ride. So what happened is they saw our development, and there's a couple of sensitive slope areas out of 150 acres, about four acres up.

Clayton Hepler (11:33)
my gosh.

Justin Piche (11:49)
is this kind of like sensitive slope area where drainage could run off into a some sort of basin of a river. Although there's no river for a very long time, but you know, it goes gradually down and eventually we'll get to some tributary that gets to some stream that gets to some river. And so they care about that stuff. They wanted us to hold that land, any of the sloped areas of the land

They want to hold it in a different entity or in an HOA or in, they didn't want us to give that or sell that land to any landowner. Which on the surface doesn't seem like that big a deal. It's like four acres out of 150, except for each lot is like 200 something thousand dollars for like a three, two to four acre lot. And so we're talking about.

by making us not give this land to people. We're talking about like a $200,000 $350,000 reduction in the gross profit of the deal, which isn't a ton, unlike the magnitude of the deal, but it's still like, there's no point in doing this. This is not how we've ever had to do it. The way other people have done it in the past, and the way our engineer had done it in the past, is to just create restrictions in the HOA and in the deed that protect that land from being developed.

Clayton Hepler (12:48)
Yeah, yes.

Justin Piche (13:10)
exclude like like show the plot shade the areas put in the buffer zones and specifically state no landowner may build anything inside of any buffered or you know protected area and They wouldn't let like they didn't want to let us do that and so it took four months of back and forth we got our attorneys involved and finally they allowed us to Pass to record the restrictions and get it moving forward

And so until then, we couldn't get a pull-up permit to start disturbing the ground to start our road. And so by that time, our contractors had to fill their schedule with other jobs, and then it took a little bit delay to get them back. And now we're about two weeks out, I think, from having our road completed. Once the road's completed, then we can do a final plat. Once final plat is completed, we can actually sell these lots. So we're approximately four months behind of where we expected to be in the schedule there. So it's a borough of Spokane.

We're a little bit ahead. We closed in October time frame and so we should be selling the first lot at the end of the month, which is exciting to move forward. And then we've got kind of our next bigger fund right now that we're raising for. We just started raising for a deal in Bastrop, Texas. So like 40 minutes outside of Austin. Yeah, it's a great deal. I'm pretty excited about it.

Clayton Hepler (14:12)
Nice.

Justin Piche (14:33)
But it's been a saga, man. These big deals, they're great in the sense that there's more money to be made on an individual deal than there is typically on a flip type of a deal. I mean, just the nature of doing larger deals. ⁓ The challenge is just all the money and time and energy and cost it takes just to prove a deal is gonna work. And then once you prove it's gonna work, then you gotta get the bank on board.

You've got to get the investors forward. And frankly, investors have not been the hardest part because these are objectively good deals and we're giving pretty attractive returns and we have a track record of doing good deals. It's not necessarily the investor funds that are the hardest. It's still always the bank that's the hardest. We've talked about this in the past, you know, how partnerships have helped me be able to do this without them. You know, it's hard to kind of do this, but we're I think we're getting to a point now with these three big deals and gosh, how much.

Clayton Hepler (15:04)
Yeah.

Justin Piche (15:33)
$10 million of like guaranteed debt on just across these three deals that like I don't know how much more even our partnership can guarantee at this point. It can guarantee a lot but you know once you start getting up into like 10, 20, 30 million dollars of debt like at what point does the bank be like okay you don't need a personal guarantee anymore or you know you gotta find banks that don't you don't need personal guarantees for which is hard in like the land space.

A lot of people struggle just to find banks in general to lend on land. There are plenty of them. Typically the lending, loan to value, et cetera, isn't as great as you'd like. Typically they're smaller, more rural banks, and so they can't handle as big of deals. They have a cutoff of what kind of deal size. But all of them so far that we've worked with want extraordinary personal guarantees, and in some cases, collect.

like other collateral pledges, et cetera. So at some point you kind of run out of, just run out of room to be able to guarantee this debt.

Hey guys, this is Justin interrupting your podcast. Say thanks for listening. We are talking deep about funds and AI and how to think about hiring in the world we live in right now. And what are some high leverage activities for you to do as a CEO of your company? If you're getting value from this, can you please rate, review, subscribe? It helps us continue to bring value every single week. And now back to your regularly scheduled programming.

Clayton Hepler (17:06)
Yeah.

Justin Piche (17:07)
So you need start thinking about institutional backing. As a no time small land investor and developer, where do you find institutional money to back you and how much you have to give up? feel like I'm always entering a new realm of trying to solve some future problem or some problem that I anticipate I'm going to have here the next six months to a year.

Clayton Hepler (17:24)
Yeah, dude.

Yeah. Well, I think I want to take the direction of the conversation and do an interesting one, right? This what this meant this I want it to be pretty raw. I want it to be interesting, but I'll give you a real time. And I think that the listeners will will really benefit from this. So as you know, everyone's very aware, you know, A.I. is

Justin Piche (17:42)
Okay.

Ahem.

Clayton Hepler (18:01)
changing the shape of what it means to conduct business. And even for me as a CEO, it's changing the way I'm thinking about hiring people.

So I think this is instantly practical as just that I have people in the deal engine every day asking me these types of questions. But normally they're coming from the perspective of ⁓ a virtual assistant global talent.

But when you are hiring an acquisition manager, now I have my thoughts around this. So I want to, I'm going to hold them back until I really hear your thoughts, because I think that the listeners will benefit from your line of thinking. But if you're hiring an acquisition manager in 2026, every year it looks a little different, right? What type of acquisition manager you needed three years ago in the land business was a VA in the Philippines, right?

And now it's around a more professional closer. Right? But the market is constantly changing and shifting. And so you need to hire the person that can handle that type of thing. So, if you were in a position that you had the pick of the litter, the market right now, the job market is not like it was.

It's very, there's a lot of opportunities for getting a really good hire right now as an employer, which is really exciting.

If you're bringing someone new on your team, it's going to be expensive as an acquisition manager. Would you have someone that is extra senior, almost like an executive level person, right? That has some experience of closing but is more organizationally strong. Would you go for a college kid that's 10%, 12 % commission, grinder, but there's a lot of

There's a lot of things that you have to work through with them or her. They're not mature in the professional sense. How would you think about making a big acquisition hire given, I'm sure a lot of the things you and I are both thinking about, which is over the next coming years, leaner teams, right? So I'm just going to, I'm leaving it purposefully open.

Justin Piche (20:34)
Yeah

Clayton Hepler (20:35)
And I think it would be a really good, with AI taking like, even lead managers, I was talking to another investor the other day who has AI lead managers. Now don't know if I'm completely bought into that, but if an AI lead manager, maybe an AI AM in the future.

Justin Piche (20:55)
my gosh, I don't even, ⁓ what's crazy is that we probably are not that far away from that type of capability within AI. It's not there yet, at least it's pretty easy to tell you're talking to an AI agent. But ⁓ back to your question on hiring acquisition managers.

Clayton Hepler (20:56)
Pretty crazy.

It is.

Justin Piche (21:21)
And I'm gonna take maybe like broaden it and just say like anybody. It's hard though, acquisition managers is a specific skill, sales skill. Like I don't necessarily need that person to have many other skills other than like high organization, high engine to call and talk to leads and follow up with leads and just incredible negotiator and person who's able to to like make build relationships very quickly with people and get things under contract. ⁓ So.

That's a great question. Dude, I honestly don't know. I'm struggling with this decision because my gut tells me, and this has always been my gut when I'm hiring, is that I like to hire hungrier, grinder-type people. People who are gonna put in a lot of effort and learn their job and really learn my business and kind of learn it as they work in it with coaching and just a lot of failure, a lot of trial and error.

and figuring things out, but also somebody who's not formed by a certain process, like a certain way of doing things. Because with this business, we all know, mean, you know, especially like, it's always changing. There's always different tactics we're taking. I can't, none of us have the luxury of just doing it one way and just continuing to do it that way, period.

for multiple years. Like we can hardly do that for six months without having to like change something about the way we're marketing or the way our contract is or whatever. always are tweaking and trying to get better and doing A-B testing. And so my worry with somebody hiring somebody who has a lot of experience and is ingrained in a certain way of thinking and doing things is that they may not have as much flexibility as somebody who I can bring in and explain from the ground up.

Because I really need them to just have that engine and those people skills. That's really what I need. I don't really care about much else. Anyway, and the fact that they're young and probably native to AI at this point is really helpful, I think. But that's not necessarily as important, I think, for an AM role. But for any kind of management role, Any kind of like technical type role where transaction coordinated, et cetera. Like those people aren't able to use AI.

to make their job better or easier or take work away from themselves, then that means you're the one who is having to do all that for them. And how much better would it be if you had freaking five employees that all could replace themselves with their jobs and then focus on higher order things, right? Keep the team lean, right? Be able to handle more volume and more stuff with less people.

Clayton Hepler (24:04)
Yeah, think it's a good thought and this is going to be present for a lot of people because it is very interesting when you hire someone, you're making this bet that by hiring this person you're going to buy back a percentage of your time to then allocate towards other activities that are higher dollar value.

Justin Piche (24:04)
Anyway, that was a long talk about...

I'm curious of what your answer is though.

Clayton Hepler (24:34)
So if I hire an AM and they make $250,000 in a year, they should make $2 million for the business, $2.5 million for the business. So if you're hiring for a very specific role, I think in 2026, yeah, it's good to have a really good salesperson, right? But other skills are also important, right?

⁓ And the question is always what is your organization need?

and can you afford to hire someone with a multitude of skills or are just looking for an and to just be closer, cold-blooded closer? Are you closing more sophisticated deals? Right, like for example, imagine if Justin, you had ⁓ five bass drops in a year, six bass drops. Right, those are more sophisticated deals. You're the one negotiating those deals, right? Ben?

Justin Piche (25:38)
Or yeah, or mean, or the partner who's bringing it to us. Yeah, that's not typically the AM.

Clayton Hepler (25:42)
Right.

So that's the, so do you want to hire someone that, this, you can see how this kind of goes and unravels, it depends on the direction of your business. And so this is something that I'm actually really struggling with personally. ⁓ Because I like the ⁓ person that can be moldable and smart, hungry, coachable, but also having someone who's fully autonomous ⁓ as an executive within your team, like your AM.

That's why I ask you this question, right? Because your AM has a base salary. ⁓

Justin Piche (26:20)
Well,

she's the ops manager, right? The AM is my original AM from the business. Who's doing it, who's awesome. I, yeah, and sorry, I don't mean like correct me saying, so she is negotiating. Yeah, yeah, yeah. She's my ops manager, but I call her, she's over, she spends about half her time on the phone with people and the other half like just managing the team and getting, you know.

Clayton Hepler (26:31)
But she, go ahead, go ahead.

Justin Piche (26:49)
giving direction for marketing, giving direction, coaching, building systems, et cetera.

Clayton Hepler (26:57)
God, she's like a director of revenue. Like front end.

Justin Piche (26:57)
Is that the... Yeah. Yeah, yeah, exactly.

Yeah.

Clayton Hepler (27:05)
Got it. Yeah.

Justin Piche (27:06)
She's

like, yeah, it's not, she used to be almost purely ops and no negotiating. ⁓ And then, and then I, and then I was just like, we you know, we need a higher quality negotiator to be able to go after some of these larger deals that we're looking at. You know, we just need our A players on the thing. And so that's why she's, her job has moved over to more on the negotiation side. And she's quite skilled, frankly, you know.

When she's on the phone, she's quite skilled.

Clayton Hepler (27:38)
Yeah, yeah. So that's the point. So that's the point. So I think it's, and most importantly, that having a business like ours, structuring compensation according to variability, I think is the best way to go. So if you have a lot of ⁓ high paying positions, six figure salaries, ⁓ then it becomes more difficult to be nimble.

Right, as if, or it becomes more difficult to go from one marketing channel to the next. That's the benefit of being smaller. You can do that. And so, it's interesting. I thought, I wanted to just present that to you. Let me tell you what I think. ⁓ So, like I said, ⁓ I'm still like, processing this.

Justin Piche (28:24)
Mm-hmm.

Clayton Hepler (28:34)
Depending on the direction of the business for me and and we are clarifying that in terms of like going after these like larger entitlement opportunities, but we're also doing some ⁓ You know standard lot on home packages someone has to be able to be incredibly versatile and understand different levels of negotiation and be able to meet people where they are at any level whatever whether that's You know a two acre lot in Murfreesboro ⁓

Tennessee or a You know 200 acre lot in Charleston, South Carolina so That's that's it man. I thought it was just interesting to just we still struggle with things you know as we were I'm constantly Trying to think through process things I don't have all the answers at all even though we have this platform and hopefully that was a helpful exercise for our listeners to understand kind of Where you and I are sitting?

Justin, ⁓ what else is on your mind,

Justin Piche (29:32)
Yeah.

still messing around with this open-claw AI stuff. And it's super interesting. Last time we talked, I was about to buy a Mac. Oh, I didn't even shut my... I have the thing. It's right here. But I'm not even gonna get it out. I have the thing. It's right here. I figured you can get it on... There's so much you can just do on the cloud. So it's like, why am I committing to hardware?

Clayton Hepler (29:47)
Yeah, yeah.

Justin Piche (30:02)
That kind of a rash decision. Anyway, I'm returning the Mac, so I'm not spending that six grand or whatever it was. honestly, I just like love the latest things. That's true, but they're coming out with a new Mac ⁓ Studio, I think, later this year. This one's the M3 ⁓ Ultra, which is essentially two M3 chips, like, put together.

Clayton Hepler (30:09)
No, it's for security, I think.

Justin Piche (30:32)
And then they have the M4 Max, which is not as good as two M3s stuck together, but it's better than one M3. And I think they're supposed to come out with the M5 Ultra, which is probably going to be designed specifically to run local LLMs moving forward. So I don't know. We'll see. I may still buy a Max Studio, but I'm having a lot of success with just a VPS terminal online, renting a space for.

six bucks a month. I mean, lot of the reason to do it was like, you can get local models on your machine for free and only run API calls to more advanced models when you need to. But in the last like two weeks, things have been crazy with like different models being open sourced and like, you know, free tier. But obviously with the free tier, you're giving them a bunch of data because they have to like, there's a reason why it's free. They're doing something with that, you know, what you're doing. But they also have a lot of cheap models.

Like Grok just, they put Grok 4.1 fast. It's like super inexpensive, like really, really cheap to run and it's quite high quality. So that's the one I'm kind of mainly running off right now on the thing. And then I, you you always have Claude Opus 4.6 or like the latest Chet GPT or whatever. And, you know, I'm not going to get into like a political discussion about

which way those models swing in terms of like their what they value. There's some pretty crazy stuff on X that like Elon Musk posted. my gosh. Like I didn't want to get into it, but essentially you can ask some of these AIs questions that are so obvious, like what the right answer is. And some of these models will, we'll just get them totally wrong because they've, they've got these, the certain, the certain moral kind of line programmed into them.

Clayton Hepler (32:02)
Really?

Justin Piche (32:24)
where they value certain people and certain groups of people and certain things far above other things, even though they should either be the same or like maybe even this other thing is far more important.

Would you, if you could prevent, have you seen the like thermonuclear war and like the miss, okay. It's pretty crazy. And so Grok is obviously much more like maximally truth seeking some of these other AIs like chat GPT or open AI and Claude, but for what I use it for, I'm not asking anything like that. Like I'm purely asking for marketing, like marketing stuff, know, marketing.

Clayton Hepler (32:41)
Give me, yeah. No, no, no, no.

Justin Piche (33:08)
prompts and data analysis and like coding and building building Python scripts to run that that can automate processes for me and it's pretty cool. I mean yesterday I literally I'm sure a lot of people have figured this stuff out before like automating their like listing process when they have a property etc. But we haven't I mean like we haven't been really working on it. We have a have a sales team that you know when we are ready to market and sell a property we have a process we have SOPs built.

We have, you know, we get our drone photos, we get the drone photos in this certain folder, then we have to like get them in the right format and downsize them, because drone photographers sometimes upload 40 megabyte photos to our thing, and it's like, my gosh, we have these giant photos, we can't upload those onto WordPress, you know? And then we fill out all these fields in Notion, and then we have some process where we get a listing description for it via these LLM prompts that Brian has built out.

And then we go to the WordPress and it's super finicky and we have this whole SOP of what information goes where. This information goes here, this information goes here. Run this prompt, do this thing, do, do, do, do, do, do, do, do, do, do, do, do, do, do, do, do, do,

Yesterday I built a, I used AI to build a script to generate a website listing just directly from notion and Google draft. Like when we're ready, as long as like the fields are there, it uses the WordPress rest API and builds a property page or brought like a property and just gets into a draft form. And it's like almost perfect. Like I'm still tweaking like these last little things. Like it has struggles with.

Sub lots like we do a subdivide the way our data is organized the notion is we have a Acquisitions card which is the property that we're buying when we're buying it right so we have we have a lead card a lead card is and this may be too technical for a lot of people but I don't know if you're if you're interested I'll keep talking if you not I'll stop You let me know you let me know what do you think keep going? Okay? So we also are like notion database is organized as we have a lead card the lead card is the seller

Clayton Hepler (35:21)
Keep going, dude. No, I think this is great.

Justin Piche (35:30)
That's the seller. So it's not a deal yet. It's just a lead, right? In Notion. And then once that lead card is under contract, it becomes a property card. Because now we have equitable interest in this property. We are controlling this property and we are spending money on this property. So it becomes a property card. And then at the same time, once we decide whether we're going to subdivide it or not, we also create a dispositions card. So we have a acquisitions card, have a dispositions card, we have a lead card.

and they're all the same deal. Now, the dispositions cards, card, there are multiple dispositions card tied to the same property card, because if we're subdividing a property into say 10 lots, we have that one property card, right, for it, and then we have 10 dispositions cards, one for each individual lot, and they're all linked back, so like, the way our financial ⁓ dashboard goes is all the costs that get allocated to these sub-lots.

either prorated for something that does the whole thing or maybe like individual cleanup or whatever, all of that comes back into the parent card. So I can track all the finances for that whole deal on this one card. What the script was doing was really only looking for the dispositions card for the lot that I'm listing and not going back one level to the parent acquisitions card essentially, and then seeing all the sublots that are related. And so it wasn't able to like fully put our like a subdivide.

break out the subdivide lot pricing on our WordPress. So that's like the last thing I am fixing with the script. Otherwise it like completely do it all. It's crazy. It saves one person probably like 15 hours a week, depending on, you know, 20 hours a week, depending on how many properties we have to list. We list five properties. That's like 20 something hours a week of time for one of my team members saved by like a half a day of work for me in perpetuity, which is

awesome. That's like, that's really the power of kind of where we are with AI and you know, anybody who could code could probably have already done that. Okay, but I can't code, right? I mean, I used to, guess in college, but like, it's not the thing that I do with my time, but with AI, but I, but I understand logic quite well and I, can explain how things should work and AI is so good now that I can take those things I'm saying and turn it into executable script. And it's so good at like error handling and everything too, right? ⁓

Clayton Hepler (37:39)
Right.

Justin Piche (37:55)
It'll give you the exact instructions on what to do. Like you run this script, you get this output. Like there's all these errors, you post it back in and it's like, oh, I think this, that, this, let me rewrite the script to handle that, And it gives you the script, you run it. It's like, here's a new error. should go, it's a lot of kind of.

Clayton Hepler (38:11)
You're like, you're

like, it's like telling you what to do. It's like, here you go. It's amazing.

Justin Piche (38:14)
⁓ yeah, yeah, yeah, I'm

not pretending to be some like incorrect. Now the logic behind how it works and like understanding why certain errors are happening because you kind of understand the file structure, you understand what's in the process or whatever is helpful. And frankly, like I don't think we'd be able to even automate half this stuff if Bryden had not spent the last like six months meticulously organizing the data and standardizing all the Dispo cards and.

Clayton Hepler (38:23)
Yeah.

Justin Piche (38:41)
creating incredibly detailed SOPs and creating AI prompts and a lot of things that have already sped up the team. So I'm able to, basically with this overclock thing, I'm able to run commands that look at all of our SOPs that are built because they're all in the Notion database and read all of them, read all of the words, read the transcripts of the Loom videos, and I'm able to give...

this agent thing access to my Zaps so can see every zap that we're using and every automation and what happens and what triggers it and where it goes and table synthesize that information and propose things to us that are like this is a task that I can like completely automate so I had to do that I had to run like through every SOP one by one and give us like give me essentially like a report of what what is the highest value things to

write scripts for and then use some minimal amount of agentic layer. The idea is like the more stuff that is Zapier type stuff, programmatic or script based that just executes step after step after step that's logic based, kind of the better because you know, agentic layer, the agent itself, there's some hallucination, there's some like things that can be wrong. It's really good, it can synthesize, it can understand, it can write whatever, but it's not

perfect. It just can't be. Maybe eventually it will be. It'll just iterate itself until it figures it out. But the idea is let's really write a bunch of scripts, Python script type stuff, programmatic things that execute these tasks and add agentic layers on top when reasoning is required or when troubleshooting is required, but not rely so much on the AI itself to run things. It's much better to of obviously programmatically do it so it doesn't have errors.

So it's able to propose these solutions and then it's kind of on us to work with it to like tease it out and then make it so it's kind of, I don't want to say foolproof, it's never foolproof, but make it so it's pretty reliable and then start testing it against edge cases, right? Because it's not, it might work, like I'll give you a perfect example. I know I'm talking a lot about this, but it's super fun, it's super exciting. It's one of the reasons why I'm staying up late is because I'm working on this stuff for the business, but.

Clayton Hepler (40:37)
Yeah.

Justin Piche (41:01)
⁓ A perfect example is I had it write a script for listing a property. And I actually used a draft property that the sales team had been working on that day. So I knew that it had all the right information and I could give it like the exact ⁓ property card in our property listing in WordPress to look at the API so it could see all the meta fields, everything that got filled out by a person. And then it can relate that back to Notion.

and kind of figure out what goes where, et cetera. So probably like a couple hours of iterating with this thing, we're writing scripts, we're putting it, we're testing it. And finally I get a result and it sucked. It sucked. It sucked. It like created it though, but it had all these errors, right? And so then I give it the upload, I show it. It actually, I'm using Claude desktop now, like with the max whatever plan. ⁓

Clayton Hepler (41:55)
Yeah.

Justin Piche (41:56)
which I'm still an Abacus user too because I like having access to all of the different agents and deep agent, et cetera, but I'm using Cloud right now. And it's awesome. mean, can control your browser. can go, can like see things and like reason with it. So that's what it did. I was like, okay, I need you to review this. So it pulled up like Brian's listing. It pulled up the listing that it had created or the agent had created. And it just like starts scrolling down on both of them and like analyzing and seeing all the fields that are missing and ever. And then it's like, okay, I see now I'm missing this, the other. And then it writes me a new script.

put it in and it finally it got a listing that was really good like it looked good except for the one thing was missing was a high quality listing description right it just it used the notion fields essentially like the data that we had in it to propose a listing description so it was terrible right it's like needs well like per zoning no counting zoning like stuff like that like in a row you know look terrible from notes or whatever on the card ⁓

But it worked. And then I ran the, like, tried to do it on another property and it totally failed. I was like, what the heck? So it's like, what was different about it? Well, for one, the photos were much larger. They were huge photos, where the other one was PNGs and they were all less than five megabytes. The other one was JPEGs and they were 40 megabytes. So then you're like, okay, well.

I don't do it. Do I want a physical manual step of a person having to go and check every photo and convert it to the right file type and compress it to the right file size? Or do I have the AI program in a photo by compressing and changing of type like step inside between downloading and Boom, that's what we did. So it was like literally like a couple of words and boom, it like downloaded it had me like download this thing. And now then it worked.

Clayton Hepler (43:41)
Yeah.

Justin Piche (43:47)
And was like, yeah, I'll do this. So it removes a step from the person. So now it doesn't matter what size or file type the photos are, as long as the right photos are in the right folder, the script is going to compress them or download them and change them to the right file type and make them smaller and then re-upload them into WordPress so that they're all there. And that in itself saves somebody 45 minutes for one listing because it just takes so long to do that. Downloading 25, 30 photos at 40 megabytes, putting each one into

into a program that converts them and compresses them and then putting them back and resaving them and then going into WordPress and uploading every single one. Like this is a ton of time repeated over and over and over again. And we're only like just now with AI and probably with our own business like systems and data and organization able to actually start to really truly automate some of this stuff that we do every single day. And the idea is like the more stuff like this we can automate this repeatable not.

customer facing stuff. The more time people can spend on the phones with people, right? The more sales we can actually do, the more follow up we can actually do. The more we can focus on like how do we sell more property faster? You know, that's kind of the, anyway. I'm pretty fired up about it, if you couldn't tell. Like, I love this stuff. This is awesome. I haven't been like this excited about like kind of the core business, you know, of like where we are and what the opportunities are in a long time.

Clayton Hepler (45:02)
Yeah, man.

Yeah.

Justin Piche (45:14)
And there's still a lot of actual acquisitions and sales headwinds, right? Just from where the market is, but the process is, I'm pretty excited about where that's headed.

Clayton Hepler (45:26)
man, it's really exciting. You know what this makes me think of is transaction coordination. Dude, it's

Justin Piche (45:34)
That is the next

step. Man.

Clayton Hepler (45:37)
Is it? You know, the one thing would be interesting is like solving the problems. And and

like solving title problems. We've had issues lately. you know what's interesting as you're talking about this, like you just have a couple of orchestra conductors on your team. Back to the conversation earlier about the AM. If you have a couple of people that are the orchestra conductors, then you can automate the majority of transaction coordination. You can automate the majority of the

Justin Piche (46:14)
sure.

Clayton Hepler (46:17)
Getting on the phones, that's the thing that remains, right? It remains. But everything else, the talking to brokers, man, that's a, that's one, unfortunately we can't automate that one.

Justin Piche (46:23)
Mm-hmm.

No, no, you can't.

And maybe eventually AI will be able to do that stuff. But I think I don't really, I'm still of the mindset that most people are not gonna prefer that. People wanna talk to people, right? They don't wanna talk to an AI.

Clayton Hepler (46:49)
You know, it's interesting. There's the,

yeah, there's the kind of the whole thing of like, well, Gen Z is, ⁓ you know, they're kind of bent towards screens and they never grew up in an environment like you and I were the last bastion of civilization, right? Of not having, I didn't have an iPhone. I didn't have an iPhone.

Justin Piche (47:09)
that didn't have screens as kids.

Oh yeah, actually dude, where is it? Oh, it's not here anymore. I was going to show you the video. We went through these old drawers that we had and my wife has all these phones from way before iPhones. And we actually pulled off, we opened one of them. Oh my god, it plugged in.

I would, this may be too much of a tangent and a funny thing, but this is from like high school. And so it has a bunch of MP3s on it, like music that she uses as her ringtone and stuff like that. it's pretty funny what songs are on there. And so we played some of them the other day. I was on my computer, clicking through playing on my daughter's there. She's like, what's that song, mom? And it's like little Wayne is, it was funny. Yeah.

Clayton Hepler (47:55)
Yeah, yeah, yeah, right, right, right.

I remember, yeah, the ringtones, right? Like,

Justin Piche (48:02)
the ringtones yeah you have them your little SD

card and you put them into your Blackberry or whatever and you can put yeah and in there I mean before that I my first cell phone was a Nokia like literal Nokia brick or Sony and then I had a Sony Ericsson kind of like brick thing and all I have a snake it's all yeah anyway ⁓ yes definitely

Clayton Hepler (48:21)
You remember, yeah, you remember razors? Yeah,

the razors.

Justin Piche (48:26)
Dude, I

had a pink razor with a, like I swapped like battery covers with somebody who had a black one or I had a black one and I swapped. It was like I had dual colored razor.

Clayton Hepler (48:41)
That's cool, man. That's cool. ⁓ Good. Yeah. So I'm thinking about like transaction coordination. You know, we're, building out some stuff just to be exciting stuff and internal sales coach, AI sales coach, ⁓ like voice based AI sales coach. ⁓ and also, ⁓ call reviews, all AI, right? Like creating a really good call review system for AI.

Not like, you know, I'm going to upload it to chat GPT, right? Like it's true. True good call review. Yeah.

Justin Piche (49:12)
Yeah, like automated kind of call

review type stuff. Yeah, those are all like things that are now like actually completely possible because what was hard before is AI probably is could have done this probably for the last year and a half, two years type of thing. You know, like you could have done something like reasonable using some sort of transcription software and then putting the transcript in and then, but now actually having an AI that can talk you through it like voice wise.

which may be more helpful for your cold caller agents, people who are like, or your AMs or whatever, because they're, especially if they're overseas, this is being more helpful for maybe a US person, it'd be easy to just read something. But even still, I kind of send them, hey, you know, when you said this, here's another way of doing it, let's role play, let's figure this out and have something that's interacting with you, which is probably way better. But also the ability of AI to understand your software stack and

help you code and actually produce code for you that works has also never really been there before. And like, I mean, FollowBoss where we do all of our sales calls, essentially, API, it's an API. You can pull everything out of there. You can see everything with these API keys. It's awesome.

Clayton Hepler (50:19)
Correct. Correct.

Yep.

It's awesome. You want to, you want to wrap up?

Justin Piche (50:42)
Yeah, we kind of went all over today, but it's been it's been it's it's interesting. I always enjoy the conversations that aren't necessarily like a direct topic, but like, what's top of mind? And right now it's obviously funds, it's AI, it's how do we optimize its hiring? What does the hiring look like in this kind of new world? I feel like I want to get to a point where the business is steady. I've always thought that it'd be nice to like get to a point where we're just like, we're like steady.

You know, we're not trying to grow, we're not trying to survive, just like, executing at a high level doesn't mean I don't want to work. I mean, I'm happy to, I enjoy work, I enjoy my work, so I like to always, but it always feels like something as big as change and constant.

Clayton Hepler (51:28)
Yeah, it does. That's business, right? Like unfortunately, we're not, what's the big thing about our business ⁓ as a ⁓ last thought is that unless you are really making, I don't know what the level is, Justin, because I haven't really gotten there, but once you have executives in every one of your departments, let's just say, ⁓

You have a director of sales, have a director of marketing, you have a director of ⁓ fulfillment, which would be transactions, dispositions, finance, director of finance, whatever. Then you can kind of take that step back, right? And it doesn't always have to be like always on your mind, which is always on my mind.

And a lot of investors to have a tendency, and I'm not saying this is you, but to like try to get back in the cookie jar, they'll hire someone and they're back in the cookie jar, right? Because they're so close.

Justin Piche (52:30)
I mean, do

feel that a little bit, you know what mean? Like especially when things are slow. I have this like almost, I almost can't stop it. It's like jumping back in, but.

Clayton Hepler (52:36)
Yes.

Yep.

Justin Piche (52:45)
But I found what's more effective is like tackling problems for people and not necessarily like the same same role you always have as a CEO, which is to make your team more efficient, to break down barriers so your team can achieve more, to get the right people in the right seats or equip them correctly or train them correctly. Like those same responsibilities are kind of I feel like they're translating into being the guy on the team that's working the hardest on developing these AI.

Clayton Hepler (53:01)
you

Justin Piche (53:12)
and programmatic solutions to a lot of repeated tasks because it's going to help the team. they have too much going on, honestly, in their roles to still be effective in what they're doing and step out and do this kind of higher order, bigger picture type work.

Clayton Hepler (53:33)
Yeah, like imagine you ask a baseball player to to I want you to practice your pitch or I want you to practice pitching. But also I want you to study all these nutrition things that you need to do in order to be perfect at pitching. And I also want you to study all the, you know, best workout right regimes. No, they have to they have. Right. Like here's the problem. You need to get stronger. You need to get stronger. Right. You need right. Like

Justin Piche (53:56)
Yeah, no. They have a trainer. You have a nutritionist. Yeah, yeah.

Clayton Hepler (54:03)
And so that's, I think that's an excellent way to end man because you take it as a responsibility man. This is my company and I'm here to serve my people. And the better they can win, whether it's me supporting them through coaching, training, mentoring, or me solving problems, that's a high leverage activity.

So, as always, gentlemen's agreement of the podcast, if you like it, guys, if you get benefit from this, rate, review, and subscribe. Justin and I, have you ever seen that? Have you ever seen that meme? Have you ever seen the meme of like the Saturday Night Live meme of like someone calling someone and they're like, well actually, you're a really nice guy and then there's another guy on the other line and he's like, well thank you.

and he hangs up the phone and it's like him in his room. And then there's like on a wall, there's like a name of people like who to kill or whatever. And I have my, the listeners, all of our podcast listeners, who to remind to subscribe to the podcast. And I don't know how long that is, but if I, you gotta see the meme, look for the meme, look for the meme. That's the who to subscribe, ⁓

Justin Piche (55:16)
you

Clayton Hepler (55:22)
⁓ Guys, I want to mark a lot of names off the list this week. So, subscribe, rate, and review the podcast. We always appreciate you. Thank you.

Justin Piche (55:33)
I have nothing else. See you guys next week.