EP.26 / Jake Heller
The AI Tools Reshaping Real Estate Development
For most real estate professionals, using AI might look like asking ChatGPT to tidy up grammar on an email, but that’s just scratching the surface of what’s possible. Jake Heller, co-founder of AI for CRE Collective, argues that the true utility of these tools lies in automating tedious tasks, freeing up valuable resources to build relationships and strategies that actually get deals across the finish line.
Jake is a third-generation builder navigating the complex realities of Los Angeles development. He’s experienced firsthand how AI tools have become a necessary partner for developers as expenses rise and rental growth stagnates. Crucially, we discuss the friction of AI adoption within large institutions and how successful integration doesn't mean forcing teams to learn entirely new software, but rather deploying tools that seamlessly fit into the workflows we already rely on.
We also touch on the nuanced regulatory landscape of LA and how Jake is utilizing specific machine learning models to mitigate inefficiencies that usually kill deals. He details his personal toolkit, walking us through platforms like Gamma for instant investor decks and Endex AI that streamlines institutional-level underwriting.
Listen on Apple Podcasts | Spotify
About Jake Heller
Jake Heller is a third-generation homebuilder and entrepreneur tackling Southern California's housing crisis. As co-founder of a development company, he built apartments across LA before co-founding JDJ Consulting Group, a leading land use consulting and permit expediting firm that has facilitated approvals for approximately 6,000 housing units throughout Southern California.
Jake also founded the AI for CRE Collective, a 400+ member community where commercial real estate professionals share AI tools and workflows that are transforming how deals get done. The collective is the central hub of where AI/technology meet Real Estate.
Topics Covered
Creating a hub for AI and technology in real estate
LADWP, offsite improvements, and the "LA Regulatory Risk"
AI essentials: Gamma, Claude, Endex AI, and Shortcut AI
The implementation gap and the reality of data infrastructure
Why prompt engineering is the new essential skill for analysts
Will General Contractors benefit the most from AI?
The roadmap for non-technical professionals to get started
Resources Mentioned
Gamma - AI Deck Creator
Endex AI - Excel AI Super Agent
Shortcut AI - Excel AI Super Agent
Claude - Large Language Model
ChatGPT - Large Language Model
Perplexity - AI Research Tool
Manus - Autonomous AI Agent
-
Sam Pepper 00:00
My guest today is Jake Heller, someone I've known for a couple of years. We met when we actually did a tour of some of the properties he was developing on the east side of Los Angeles a couple of years ago, and then since then, we've kept in contact, and he's now developing a really interesting community that helps folks in real estate understand AI and the tools that can help their business and also advance the entire industry. I don't think I'm overstating that. So excited to have him on the show to talk about what he's up to and to share a little bit about the details of AI and some of his knowledge he's learned over the past months, really doing a deep dive into it. Jake, welcome to Building. La, thanks for having me. So I'm really interested, first and foremost, what drew you from development into AI, you're now doing this really deep dive, setting up this community. I've seen all the videos you're producing online. I mean, you're clearly spending a lot of time in this space. What led you to get started in that?
Jake Heller 01:05
I've always been pretty forward thinking about technology and AI and ChatGPT was released, Claude was released. I was all over it. It became a part of my life, very quickly, a big part of my life. But the AI for CRD Collective, which is our community happened by accident, and some people don't believe me when I say that, but I had no expectations or no intentions whatsoever to take it to what it is today. My partner in the collective who I started this with was initially pushing us to start a SaaS company build software, which I had no interest whatsoever in doing. I don't know anything about engineering, not even sure where to start with all that huge bandwidth issue. Just didn't have the time nor resources to pursue something like that. So what I proposed instead was, let's bring together folks who are thinking about this stuff the same way that we are. We believe in strength, in numbers, and maybe there's some ideas, maybe there's some things that we don't know about at the moment that we could then take and utilize in our respective businesses. He's on the brokerage side. I'm on the development side. And one thing led to another. We created a community, and the first day, I kid you not, we probably had 100 people sign up, and we just couldn't believe it. We were in total shock. And so we realized that we had something good on our hands, and that clearly people are thinking about this stuff the same way that we are. And so we started running with it. Fast forward to today. I think we got about 430 people. We're growing week over week. Our members, who are part of the collective love it. We've only received positive feedback so far, and I'm just a sponge at this point. I mean, I'm having conversation with so many different people. Founders in this space brokers have never done a deal before, but want to learn this stuff. We're doing structured learning academies. I've presented at a few extension classes and all this stuff. So it's been really fun. But when I tell you it happened by accident, I really mean, it did.
Sam Pepper 03:07
Did you know when you were starting it that AI was going to reshape our industry? Or is that realizations that have happened over time as you've learned more going through this process, because, I mean, we're all learning next to each other, right? Some people are more advanced than others, such as yourself, but this is new technology for everyone.
Jake Heller 03:29
Yeah, no doubt about it. Look, I have made it my job essentially to know AI in our industry, at least for the last few months. That is before that was just kind of scratching the surface of it, but there's still so much that I don't know, and I know enough to be dangerous. But I am by no means a technical expert. I don't have a coding background. I can't really speak to the back end or technology aspect of this, but I do think that AI is without a doubt, going to change the way that we do business in all different areas, and it's changing so rapidly, almost every second there's a new feature, product, tool announcement that changes how I personally do things, and my goal is to surface a lot of the use cases and new tools and features for everybody, because this stuff is so overwhelming at the end of the day, even I'm overwhelmed, and I do this all day long. So I can only imagine the folks that aren't in my position and that are kind of just going about their days. So But to answer your question, yes, I think we are only in the beginning phases of all of this. There is so much more to come. I mean, we're not even at AGI or super intelligence yet.
Sam Pepper 04:45
No, this is something I read recently, but we are what we're experiencing right now, which is blowing everyone's minds, is the worst version of AI that we're going to experience, which is kind of crazy to think about. Well, we're going to talk a lot about AI and the collective. Created on this, before we jump into that. I also want to give our audience a bit of a background into who you are and your kind of background in real estate. Again, when we met, you had co founded, I believe, MIDOS development. You were also starting JDJ Consulting, and so you had, at that time, two years ago, still, I mean, a lot of irons in the fire. And you were, you know, putting money up and building projects on the on the east side, so zooming, sort of back out and kind of back to a few years ago. What were you working on? Maybe share with our audience some of the projects that you've developed in Los Angeles.
Jake Heller 05:39
Yeah, so I'm a third generation home builder. I don't actually build homes, but I've been around it my whole life. We were private home builders the last 30 years, Nevada, Arizona, California. It's just what I've known my whole life. After the GFC, we kind of morphed into more of a family office and less of a production home builder. Didn't really start the machine back up again. It's just too hard to compete with the big boys nowadays, all the publicly traded home builders, they just have so much scale. But family's been developing. It was always interesting to me. I have cousins and a lot of family that work directly with the family office. I've always wanted to work in parallel and kind of do my own thing. And so after school, I've just been entrepreneurial by nature, and fortunately been able to throw a bunch of stuff at the wall see what sticks. So I partnered with some contractor friends of mine, who were building for others as a third party GC, and they wanted to start doing development. Not that I was an expert in it by any means, but I kind of had that background, and so we got together and started putting deals together.
Jake Heller 06:44
We did a handful here in LA nothing fancy or big by any means, small sort of standard stuff. Biggest was up to, I think, was 43 units. We did a 25 unit, a 26 unit, a lot of it in that kind of East LA market, Silver Lake Echo Park, Koreatown, and was a lot of fun. We made a lot of mistakes. We made a lot of good people along the way. We had construction capabilities in house, though, which which was nice. And while that was going on, we also started an entitlements business, which is still operational today, does about seven figures in revenue. Got a full team. Our kind of sweet spot is your traditional multifamily projects in Southern California, with an emphasis on LA City, where we have most of our relationships. But there's kind of two buckets in that business. There's the actual entitlement side, which is primarily dealing with city planning, and then there's the permit expediting side, where you are liaisoning with La DBS and Building and Safety Department to get their project RTI ready to issue.
Sam Pepper 07:49
So that consulting business is your client base and seven fixed revenues, nothing to sniff at. So is your client base kind of, I'm assuming, a sub institutional level of folks, family offices, things like that, that are also maybe developing in LA for the first time and are overwhelmed by the Byzantine kind of nature of City Planning here and then the permitting process.
Jake Heller 08:10
Yeah, pretty much we never really work with institutions. One reason, they have their go to guys and gals, but oftentimes it's a land use attorney, which we are certainly not, and we don't claim to be, so not even quasi institutional. I We have done bigger projects. I think the biggest we did was about 300 units. Where we got really active was in the affordable housing space when Ed one took shape in LA cities. For those that don't know what that is, it's executive directive one. It was a streamlined approval process here in LA for 100% affordable projects, no LIHTC, you know, any sort of tax credit deals by any means. It was just private affordable development, which a lot of people started to do just because there was a bit of flexibility with the city. They basically said you could do whatever you want in the beginning, that certainly was the case, but then they started getting more and more restrictive with this program. And fast forward to today. There's very few to no. 81 projects whatsoever, and ones that did get approved haven't really got out of the ground, unfortunately. So the city did find a way to encourage folks to build affordable housing and realize that it was working so well, they they stopped allowing it.
Sam Pepper 09:25
I'm curious. So when you're developing these projects that are, you say small, but I think a lot of people would say, you know, 30, 40, units is not small. It's all relative, right? I feel like I have a conversations with people all the time that are in the real estate industry, and they see the opportunity in LA to build multifamily housing. They have some friends and family capital, or they just are really good at hustling, and they want to do that 40 unit apartment building. That's their goal. Pulling it off is another story, right? Actually getting it built. So I'm curious kind of what. You found to be the hardest part of building those projects on the east side. Was it securing the land? Was it the construction? What surprised you, what stressed you out, what were the moments where you kind of had your heart in your mouth?
Jake Heller 10:16
What I would say, generally speaking, is that developing real estate, specifically apartments in Los Angeles is very, very difficult. Every aspect of the process is difficult, securing the land right at a basis attractive enough to work, going through the process of entitlements and permitting, dealing with the LA Department of Water and Power, which I can't even begin to describe, how backwards that department is. They make your life so difficult, and anything goes. You could be on the one yard line about to get into the end zone, and all of a sudden they say, oh, you know what? We changed our mind, and you're actually going to need to do this, and it's going to cost another million dollars. They don't care about the project or any of that stuff, right? Like for them, they have one objective, and that's just to satisfy themselves.
Sam Pepper 11:09
And they've got so much power to the folks of the inspectors, that's no exaggeration to say that an inspector can single handedly add a million dollars to a project without really thinking twice about it.
Jake Heller 11:22
Totally. And I had someone we did consulting work for yesterday call me and said, Hey, I just got a letter from LA DWP. I need to put in a fire hydrant, and it's going to cost $100,000 beyond that, they didn't have enough water pressure in their main line, and so they need a fire pump. And all these other considerations, which, like nobody else cares about besides you, because you're the one building, the building and your investors. Of course.
Sam Pepper 11:45
I mean that example is also fundamentally crazy, that the city puts that on the developer who happens to be developing on the parcel that's adjacent to where the where you have that need, and suddenly it's the developer's responsibility to build that fire hydrant. I think a lot of people will be surprised that that's the case, but that is the reality, right? These offsite improvements and can really spiral and get you in a world of hurt.
Jake Heller 12:05
Oh, yeah. I mean, on one of our projects, we had that same water pressure issue. We were at the top of a culdesac, and there was, I think there was only a couple of options, one being that we had, essentially, we could have put in a tank in the ground, which was the size of a swimming pool. And this was a 25 unit apartment building.
Jake Heller 12:18
We toured this one, right? This was the one that kind of looking over the highway on the east side.
Jake Heller 12:28
Yes, that was one option which obviously not feasible to the project by any means. Or we had to upgrade the main line, the main water line, for LED WP, and this doesn't account either for the transformer issue that we had on the site as well. There was a power line which turned out to be in Caltrans jurisdiction. We had our plans and everything approved with LEDWP, not once did anybody talk about Caltrans. All of a sudden, we're starting the project and get going on this stuff, LADWP comes back to us and says, Hey, this poll is on Caltrans jurisdiction. You either need an encroachment permit or we're going to need to do an align extension a half mile away, and that's going to cost you however much money. So LADWP is absolutely without a doubt, one of the more difficult organizations you're going to need to deal with. And I could tell you as well that the city of LA and all the different departments associated with it, including the housing department, don't make your life any easier either. So regulatory, I would say, in LA, is a huge reason why so many institutions, why so many developers, have the city blacklisted for development.
Sam Pepper 13:40
The one part of our industry that feels like they could really benefit from it is the city permitting process. If this podcast has had any theme, any sort of through line, every single person who has built anything in LA, no matter if it's a ADU or it's a 400 unit apartment building, has, at some point, had been thrown an incredibly unnecessary curveball from the city at some point. And it's remarkable that for a city that every speech that a mayor does and council members do, housing is always the first thing that's always the priority. This is the priority, but the reality on the ground is building anything, including affordable housing in LA is no easier now than it really was four years ago. Maybe some hurdles have been taken down, but they're kind of replaced by other hurdles. And quite frankly, it's a bit of luck.
Sam Pepper 14:27
At the end of the day, you can you can hire the best land use attorney. You can hire JDJ to help you with your entitlements and everything and expediting. At the end of the day, you can be doing final inspections and suddenly realize that you've got a major problem on your hands, and it's happened on our all our projects. The challenge is that on the projects that Lincoln does typically, not always, but typically they are large. Large projects, and we have a big team was incredibly experienced, and so we make sure that we have we're planning for all of that.
Sam Pepper 15:07
But the budgets in these projects are pretty large because they're institutional scale, right? I mean, our lenders are they could be pension funds for someone who wants to build housing for the wall that's building 40 units your budget is tighter than that. It doesn't have quite as much contingency potentially, and it can't absorb the cost of a new transformer or relocating a power pole to some accounts. So it's there's just a lack of understanding among all of the city governments about, actually, on the ground, how difficult it is. And I find that to be incredibly frustrating, because it seems like it's something that our industry talks about all the time, but then nothing happens at all.
Jake Heller 15:52
Yeah, there's regulatory risk, and then there's LA regulatory risk. They're not the same, not even close, and you can't underwrite that. You cannot underwrite that, that risk.
Sam Pepper 15:57
All right, so moving over to AI and kind of what you're working on now, what your focus is on now, was there another platform you were aware of that was focusing on educating our industry on AI? Because I certainly wasn't.
Jake Heller 16:11
Not at the scale we are currently doing. There were others in the industry, on LinkedIn, on other socials, talking about this stuff, I want to say that we were the first community that was focused on AI for commercial real estate. I'm not 100% certain about that. I know there were others that launched almost immediately after us, but I'd like to think that we were kind of pioneers in sort of the educational AI for CRE type work.
Sam Pepper 16:42
I mean, now that you have more than 400 members who's joining the community from our industry, is it all the brokers, developers, architects, who are the most active people in the community.
Jake Heller 16:53
So it's a good mix of people. I mean, you have brokers who have only done a handful of deals and are just trying to learn about this stuff. Have multi billion dollar developers. We have wealth managers that have real estate allocations and invest in real estate. There's software engineers in the real estate space. There's founders building prop tech and AI in the community. So it's really all across the board. I'd say there's there's definitely a good amount of real estate brokers, there's a good amount of real estate principals, there's lenders, there's appraisers, there's people all over the place. And I'd say there's a good mix as well who are active in the community.
Jake Heller 17:37
I'm certainly trying to get more engagement, but there's been a handful of people who have been good resources for those that are a little bit more elementary in their AI capabilities, which is nice. It's not just me kind of telling everybody, Hey, this that one another, there's there's other people in the community, other members who are even more technical and capable than than I am myself. And so they're a great resource to have in there, but it's really what I'm trying to create. What we're trying to create is an abundance mentality. We want to give we want to share what works. We want to surface the winners. We want to talk about the losers, right? So people could avoid that mistake by using a tool that's maybe not that great. So it's all about an abundance mentality. We really, really want to curate collaboration in there and all about giving before taking.
Sam Pepper 18:26
Yeah, so I joined the collective probably a couple of months ago, and I sort of taken it upon myself to try to educate myself about AI while at Lincoln and to make sure that A) we weren't falling behind our competitors and as an individual, right, not just as a company. I'm, I think, very aware of the fact that five years from now, if you aren't using AI as effectively your teammate on anything, you're going to be falling behind your competitors. But I had naively thought that I would be able to jump into it easily and understand sort of the field of options out there, and be able to utilize the right tools the right time.
Sam Pepper 19:12
But coming into the community, I was actually amazed the number of different applications, software platforms out there that have been launched that are specifically targeting our industry. And it took me a moment to really recognize how much I had to learn and how important this community is to assisting that. Because it's so easy to say, Oh yeah, I'll use AI, and of course, I've downloaded chat, GPT and Gemini, I'm using that, but those are only scratching the surface, and all these other platforms that have been launched, you're diving into. So I'm curious what you would say has maybe surprised you the most about the questions you've encountered from folks. What. Do you feel like people want to solve the most with AI?
Jake Heller 20:02
Well, I think that there's been a big mindset shift in the last couple of years in real estate. When you look back to just a couple of years ago, you had very favorable market conditions. Cap rates were compressed. Interest rates were lower maybe three years. But people are making money hand over fist, and when that's happening, you don't want to change what's already working. If it's not broken, don't fix it. So I think the distinction between wants and needs as it relates to the mindset is important. A few years ago, people wanted to know this stuff, they wanted to know technology, they wanted to know AI. They didn't need to know it. They didn't need to implement it.
Jake Heller 20:43
And if you fast forward to today, now you don't have the same tailwinds you did in the market. In fact, there's substantial headwinds. And so people in the space, you want to create value. You want to increase value of your assets. You want to create value for your investors. You want to increase the bottom line, increase, your noi, and there's not a lot of ways to do that at the moment. There's very little to no rent growth. In fact, there's declining rent growth. Expenses are going up. Inflation's going up. There's there's a lot of headwinds. And so I think there's fewer ways to create value nowadays, but technology does create possibilities for finding and creating value and finding efficiencies. And so the mindset, I think, has changed from a want to a need. You see competitors in your space using AI, you see industry peers using AI, you see the broker closing 20 more deals than you, and they're using AI. And so now you need to know what's going on, I think, and you need to start thinking about implementation.
Jake Heller 21:43
So there has been a huge mindset shift, I think, in the last year or so, and there's a lot of tailwinds now for AI in our space, I think real estate's the perfect industry at the moment for adoption of this type of technology. It's huge. Real estate is such a massive industry, and it's just been so slow to adopt this kind of technology. But I do think we're on the brink of mass adoption in the space, and still today, there's not enough people talking about this stuff. In fact, I was presenting at a Uli master class last night. It was 20 people ish, and I brought up Claude, which is one of the large language models, and I had somebody say, What's Claude? Claude is arguably as big as chat. GBT, you know. So what I've realized is that I'm in a bubble. You're in a bubble in the sense that there's a lot more people outside this bubble that are not familiar with these types of tools, that are not familiar with the type of efficiencies they could find by using some of these AI tools and resources. It's been interesting to see that. It's certainly been surprising as I'm having conversations just with how little people know about all this stuff. And I think there's a big difference between knowing these types of tools and knowing what's going on and then actually implementing it.
Sam Pepper 23:06
What are you using right now in terms of your day to day?
Jake Heller 23:11
There's a few tools that I'm really excited about and that are part of my arsenal. Gamma is a deck creator, an AI deck creator, which, if you haven't used it, is quite brilliant. You could watch in real time it create a very professional deck in minutes, if not less up to 50 pages even right? So that's had a lot of use cases for me, specifically, some that come to mind are generating OMS, generating bovs brokers opinion of value, generating investor decks. As we're talking to investors, family offices and folks about the types of projects that we're doing, we've been creating investor specific decks for every single investor that we talk to, which ordinarily would take a very long time to create, but we could do very quickly now. And so if we're talking with X Y and Z family office about our project, we'll create a deck X Y and Z family office X Westmont partners, here's why it's the perfect collaboration, and here's a nice, beautiful deck.
Jake Heller 24:17
And I think they really appreciate that it allows us to be different, and so that's been a really good use case for me. Is utilizing gamma. There's a ton of other tools as well. In the last week alone, though, I've been spending a lot of time with a couple of Excel AI super agents. What I mean by that are AI agents that live within Excel.
Jake Heller 24:39
There's a company called endex, e, n, d, e, x, a, i, and then there's a company called shortcut AI. And this is more of a detailed conversation, which we could probably get to but my whole theory about underwriting, specifically in real estate, is that you want tools that integrate seamlessly with your existing workflows. There's been a lot of tools and software's out there trying to tackle the underwriting space. They haven't really gotten the traction. I think that's due to a couple of different reasons, but number one is in this is relevant for you, being at Lincoln, but Lincoln's a long term operator. They've been around a long time. I'm sure they have models and proformas that they've iterated over the last 20, 3040, years. And so to ask someone like Lincoln or another group that that's what they've come to know, and it works for them to abandon that essentially, and then come on to your interface and just start using your tool, ain't going to happen. So I think Excel is here to stay for the time being. And what makes endex and shortcut so interesting is that it multiplies your capabilities without replacing what you're already doing.
Sam Pepper 25:56
Help us understand that a little bit more. Help me understand that a little more if you have an existing model in Excel, is it able to automate certain functions in that? Or what is it actually doing?
Jake Heller 26:11
Way more than that. It's essentially like a PhD level analyst as far as its Excel capabilities. So you can build proformas from scratch that are institutional style, like development models, 15 different tabs, unbelievable Excel formulas and functions, none of which I'd be able to do on my own. You know, your if errors. And so I think on that point, it lowers the barrier to entry for a lot of people. Or maybe starting to look at other asset classes, because ordinarily, as it relates to underwriting, in these models, you know, you go online, you do homework, you maybe buy a template that template doesn't really fit within your needs, and you're changing it, and you're iterating it.
Jake Heller 26:53
But if you could build like a model yourself, done with AI, that's expert level, that lowers the barrier to entry, but beyond just building proformas. I mean, you could do that stuff within Claude, within chatgpt, perplexity, great models at that. But you could have your underwriting template. And this is an Excel plugin, these AI agents. And you could, let's just give an example. You could take an offering memorandum, a PDF. Let's say it's 40 pages. You can drag and drop into your Excel model, and you can just say, update my model for me and underwrite this deal. And it'll do that without you doing anything.
Sam Pepper 27:30
It's leveling the playing field. I think AI is going to do that quite a bit, because the resources to I mean learning fairly complex things and much beyond real estate is now accessible to way more people because of these online platforms we've always seen. And this is really interesting, because I didn't know about that new plugin, AI is incredible at giving you making essentially a back of the envelope analysis, just a little bit speedier than it already would be, but it already is back in the envelope, because it's supposed to be quick. And so we've certainly seen that in terms of deals coming through the door.
Sam Pepper 28:12
And, you know, I've certainly looked at this, it's good at kind of eliminating, what are the duds, right, or where, where you need to get to on a purchase price in order for it to really makes sense. And then, then, you know, if it's worth looking at further, then you put it into your actual pro forma and run some analysis. But you're absolutely right. Everyone has their own way of doing things. I mean, capital partners, our capital partners, have their own way of doing things. We have to often fit within their kind of system based on how they want to see a model, but we're never going to utilize a underwriting platform that changes that substantially, it's not going to happen
Jake Heller 28:49
Exactly, and there's a lot of other groups like that too, but I think you hit you said something which resonated with me, and it does level the playing field. It allows people to analyze and understand data on a level they couldn't ordinarily do. And I'll just give you a quick example. We're doing a deal with DR Horton, the biggest home builder in the entire United States. As you can imagine, it's very bureaucratic, it's very institutional, and the way they structure deals is unlike anything I've ever seen before, and so my model isn't really equipped to handle that type of structure.
Jake Heller 29:25
I either need to hire an expert analyst to go ahead and underwrite it for me, or I could build exactly what I need utilizing one of these Excel agents. And so I did just that, and we've had a couple of different deal structures and have gone back and forth with them, and the way that their capital inflows and outflows are done is just, it's not even math. I mean, it's so complicated. And so for me to be able to build a model and then look at that data from a comparison approach and see, okay, which one of these is. Actually better what's the most efficient way of recycling capital and being able to understand that it's changed the game.
Sam Pepper 30:06
In that example, are you open about the fact that you're using AI? Because I think AI is often right now, something that everyone's using, but no one wants to admit they're using right because in some ways, at least today, it's like, if you you know if someone, or a student, right, writes a really good paper, but it's actually written by AI, suddenly it doesn't really count anymore. If you build a deck, but you didn't actually build the deck, the AI build a deck. Do you think that it lowers the value of what you're creating?
Jake Heller 30:36
Yeah, I'm certainly not. Why for me, I've gotten comfortable with the idea that that we're heading into this future where AI plays, you know, an instrumental role in how we do business and how we how we do things. I've accepted it, and those that haven't, that's totally okay. Everyone is entitled to do their own things. That's totally fine. But for me, I've gotten comfortable with using it, with admitting that I'm using it, and as it relates to real estate, if I'm doing this deal with DR Horton, there's substantial capital at risk. And if AI allows me to be more educated, to be sharper and to ultimately figure out the best path forward, it's a no brainer, in my opinion.
Sam Pepper 31:20
You mentioned underwriting, obviously, and underwriting being one of the things that I think is AI sort of struggled to provide a good solution for but maybe that's changing. What do you think in our industry AI is very, very good at right now, and what it's maybe not so good at within the real estate industry?
Jake Heller 31:37
Well, let's start with the vegetables. First. What In other words, not so good at, like I said previously, there's a big difference between knowing and actually implementing this. And implementing AI at the moment is very difficult. You need to have an AI automation expert or a full team or a ton of money to effectively implement especially for a group like you guys, Lincoln, right? That's an institutional and I think a lot of the issues lie with the data at the moment, and data infrastructure and being able to consolidate it and organize it and effectively communicate with it and understand it in ways that that wasn't possible previously. So I think it all starts with that, and a lot of that's happening. I feel like institutions, larger shops at the moment, are trying to understand the data before implementation, which makes sense. It's a full top down approach.
Jake Heller 32:28
I do think next year 2026, is going to be the year of implementation. But at the moment, knowing is what's happening. A lot of people, myself included, we know these tools. We know how to use them, but it's still kind of fragmented in that sense, if that makes sense, right? So sure I can create a deck, but there's a lot of manual parts to actually getting that going. There's things that happen before that run and so something, and this is specific to underwriting, but I think also applicable to the to the general theme here, but underwriting previously, it was very fragmented. You had a model, you'd get an OM, maybe it's a PDF, then you get a rent roll in Excel, and then a PNL and a PDF, and you're spending a lot of time extracting that information. You do a pivot table, and it's all about data entry. At that point, you're not even analyzing the deal. You're just doing data entry. And then you go on a costar.
Sam Pepper 33:21
You're on version 48 pretty quickly, yeah.
Jake Heller 33:24
And then then you're going to another interface right on costar, and you're looking for comps, and then you're on apartments.com and then Zillow, and then you're here and there, and you're all over the place, and you're piecing together this deal in a very inefficient way, but that was the reality just a very short time ago. I think now we're in this tweener phase, where a lot of it could be automated. In fact, there's tools out there that you upload your OM and rent roll in Excel. They could be in different formats, and it'll automatically map it into your Excel file. There's tools like endex and shortcut, which you could do that very easily. You could do a lot more than that. But where we're heading, I think, is a lot closer to 100% automated.
Jake Heller 34:03
Now I never think that things will be 100% automated with AI, and that kind of leads me to the question a lot of people are asking, is AI going to replace me? No, it's it's not. Humans will always have judgment and storytelling abilities, which AI will not and so I think AI is here to multiply our capabilities, in other words, with endex and shortcut. As it relates to underwriting, if my team could look at 10 times the amount of deals using AI, wonderful. It's multiplied our capabilities. It's not replacing my analyst. I think the best analyst, the rock star analyst, in this next wave of underwriting with AI, will be those with the best judgment. Will be those who understand what tools to use, what to tell AI, what to dictate AI, how to prompt it effectively.
Sam Pepper 34:53
I think our role is going to be front end and back end, right? So I think the term prompt engineering is probably going to become pretty commonplace, right in our society, where and more so, right? So where there's actually going to be anything in education around how to prompt AI to then do the job that you want it to do, because you can quite easily get that wrong right now and then have the AI sort of hallucinate or provide you bad data. The other thing is, then us humans need to also guide and interpret and evaluate. And I think one of the things I've seen is that folks sometimes are really excited about AI, they kind of copy and paste and that you can get into trouble pretty quickly by just using the AI as an efficiency tool, but not actually interpreting it, not adapting it, not checking it. And that is going to be also, I think, where folks are going to be incredibly useful.
Sam Pepper 35:58
The other kind of big thing here is, and you kind of touched on this, the evaluation this and in real estate, like the the assessment of risk, that's still going to be a human element. And I think the one thing that AI will never have, I don't think, is the ability to create trust. AI is just going to give you the facts, and it's going to give you more facts than maybe your intern and certainly maybe your colleague would, but conveying that to someone else, building a relationship, building that trust, that's going to still be a human component that AI just won't be able to touch. So I agree with you. I don't think it's replacing our jobs. I think it's gonna level the playing field. I think it's gonna make things more competitive, potentially, it's going to bring people into our industry that weren't in the industry, which I think generally is a good thing. And potentially broader, you know, it's pretty good for just the overall economy to have people who, maybe, who didn't have a shot now have access to this incredible learning tool that are able to get pretty efficient and pretty proficient. Excuse me, at some skills that before they wouldn't have access to.
Jake Heller 37:04
I 100% agree. I think it'll allow teams to operate a bit leaner, to be more efficient. It's going to be exciting to see all this, but, but AI agents, you know, automations, these types of things are being talked about quite a bit, not actually being implemented. And although in some cases they are being implemented, it's still quite a bit broken. So I think we're close. We're maybe at the five yard line at the moment, but there's still a lot of technical expertise and technical implementation that needs to happen for this to be adopted in a very efficient and practical way. In other words, for a group like you guys to effectively implement AI, it needs to be a top down approach. It needs to be comprehensive. It needs to be very strategic. And at the moment, that is expensive and that is technical, but this stuff is moving so fast by the second. And I think another interesting kind of theory I have is that software is dying. You have all these different interfaces, all these companies building software, and I don't want to learn another software, I don't want to learn another interface, I don't want to get onboarded to another tool. There's enough of those at the moment. So I think a lot of folks who are maybe listening, or that will listen, if you're building in the space, build something that integrates seamlessly with what we're already doing. Those are going to be the winners,
Sam Pepper 38:20
When we're looking at the disruption that could happen in real estate. There's different roles, really, there's so many different professions. I mean, one of the, one of the great things about real estate, there's so many different people that can participate in real estate, right? You've got the architects, you've got marketing, you've got brokerage, development, construction, property management. My feeling is that in terms of who is most likely to benefit from this, but it could be the contractors. And the reason I say that is because at the end of the day, the one role that you aren't going to be able to get rid of is construction. You are still going to need a team to build it.
Sam Pepper 39:04
If contractors can start expanding their role into kind of overlapping with architecture, overlapping with development, learning those skills through AI, developing tools that can create concept designs, which AI can do pretty well right now they have, I think, a really interesting future. So I would put a bet on particularly the big boys like Turner Clark or those guys, if they are able to adapt really quickly, they could be in position where they can start dominating the industry. And I think it's interesting, because we've always thought about, oh, well, modular construction is going to take over, and it just never has, because at the end of the day, you still need a team building something. But if they can expand that and get smarter on other aspects the industry, I think developers and architects could be in. To have a problem in the future. What are your thoughts on that?
Jake Heller 40:02
Yeah, it's something I haven't thought a lot about, to be honest, but I agree. I would bet that all of those companies, Turner, you mentioned all the big boys, they are spending boatloads of money on this stuff at the moment. So, I mean, you also brought up modular. It's interesting. I got roasted on LinkedIn by a home builder, which, which I know that industry well, just because that's what my family did. But this was a very large home builder. They did about 2000 homes a year, which maybe that's a billion dollars in revenue. And so I made a post about modular and how it's great, you know, it works, and it's the next wave and all this. And I had this guy comment, and he basically said, it's, you know, if it's so easy, then do it. And I was like, Well, no, this guy is about to destroy me. And so I was curious to know more, so I ended up getting him on a call, and he's telling me about homebuilding, which is a very it is tough operationally for homebuilders. I mean, it is a machine to be a production home builder. It is a machine arguably more efficient than any sort of modular building.
Jake Heller 41:05
But what he said to me was, we are doing incredibly well. We make a ton of money. And a lot of people talk about this modular and say, Oh, these home builders need to be doing it and stuff. But why on earth what I risk, what I've spent, you know, 50 to 60 years building where we make nothing but money, and start implementing this stuff, which arguably has very little utility and is just so fragmented and broken at the moment. And I thought that was really interesting. You brought up modular. But the idea that these contractors, these home builders and stuff, that's a tough nut to crack. They've been doing things a certain way for a long time. They've done it profitably. They've done it well, and they're very, very efficient. And so I'm interested to see how AI and technology will eventually penetrate that particular group, and I think you're right. When that happens, I'm sure it will, maybe a lot sooner than we think. It'll be interesting, no doubt about it.
Sam Pepper 42:08
Looking ahead, what are your hopes for the collective that you've created? Because we're still in the very early days of this.
Jake Heller 42:17
Yeah. I mean, we launched August 12, pretty much two months on the dot, and we're at 430 people. Big picture. We want to be the central hub of where AI and technology meet real estate. I believe all this is it's the gold rush at the moment with AI, everyone's jumping in. Everyone's trying to be a part of it. There's people building software. They're sifting through gold effectively. But we kind of want to be the Levi jeans in the mix right back then, they created the genes for all these guys who were sifting for gold and stuff, and they're still around today. I mean, they've done incredibly well, but we want to be the ecosystem around all this stuff. I have no interest in building software. I want to bring this all together, piece it all together. So beyond just our community, which we're on school, it's a community platform. We have a podcast, we have webinars. We want to do events in person and online. We want to have structured learning academies. We want to have databases and partnerships. We have newsletters. We want to be the ecosystem around all this stuff. So that's what we intend to do. We're talking with folks in the UK and Ireland and the Netherlands all over the world. Really about growing this brand. We intend to be a global brand. And so if someone's building software in this space, we want them to come to us. If a broker wants to learn this, anything about AI in the space, we want them to come to us. If someone's looking to hire somebody for a prop tech company or AI in the space, come to us. Look, it's a big enough pool for anyone to come in, right? Say, like we're the only shop or group that's doing this kind of stuff. But we're thinking big picture about picture about we want to be the place where all of this stuff intersects.
Sam Pepper 43:47
I equate it a little bit to sort of a Duolingo model, because you're learning a new language, and because it is so intimidating, right? And you have a Google spreadsheet that you provide people access to through through the collective and it has a list of all the platforms that are out there, some better than others, obviously, that are designed to help people in real estate. And I don't think I'm exaggerating when I say there's about probably 300 different tools out there. So it's really, it can be really intimidating to know where to start. So if someone is listening to this, and maybe they're a seasoned professional, got 20 years of experience, they want to get an edge over their competitors, and let's just say for conversations sake, they can say they're a broker. Where would you start?
Jake Heller 44:40
I'd start by joining the collective.
Sam Pepper 44:42
Pay the subscription, join the collective, and then what's step two?
Jake Heller 44:45
The reality is that this stuff requires a time investment. If you're not going to make that time investment, you won't get anywhere with this stuff. But we hope to make that process a lot easier for folks, and so when you're in the collective, what we started doing is creating role specific. Specific material. So if you're a broker, it's going to be very clear where all of our ideas, workflows, tools, automations, etc, are for brokers. If you're a developer, same thing. We've just created a prompt library, which you touched on prompt engineering previously. That is a very important thing. You get out what you put in. So we've just created this prompt library we are working with, with founders, with software groups, with Prop tech groups, exclusive, sort of member benefits, right? Free credits to their platforms, discounts. Kind of going off topic there.
Jake Heller 45:32
But as far as where I would start, I'd start by just getting involved in battle testing these different tools and sort of copying the use cases, which we do, step by step, video demos on in the collective The reality is, practice makes perfect, and if you're not going to practice this stuff, if you're not going to make the necessary time investment, it's not just going to magically show up and jump in your lap. Maybe it will. It's funny, actually sometimes, like, I always think that software catches up to you, and just recently creating Excel files and spreadsheets utilizing these, these llms, large language models like chat, GBT and Claude and perplexity, they weren't capable of doing that. To take something and have one of these large language models just create an Excel file, whether, you know, maybe it's comps, or whatever it is, right? That's a very efficient use of time.
Jake Heller 46:25
And so two weeks later, Claude then develops those capabilities. And so, like, it's like, if you're patient enough, like the software does come to you, but it won't be the case for everything. And so I think you got to make the time investment. You got to start asking the right questions. You got to start learning this stuff, because it is a freight train. It's not stopping. This stuff is not stopping. This will be a huge part of our life, and at the moment, it's very overwhelming, like I said, for me included, but the goal of the collective is to just make that a lot easier.
Sam Pepper 46:56
What do you think the skills are that will define the next generation of folks. Is it prompt engineering? Is developing relationships can be more important than developing skills in the future?
Jake Heller 47:09
The ability to know what tools to you use, how to direct AI, those are going to be very incredible skills. I feel like I know enough to be dangerous at the moment and time and time again, I'll have people ask me, like, oh, man, I got to do this. Or, hey, how would you go about this? And I honestly just do it myself, because I know I could do it in 15 seconds and it would take them three hours to do and the reason is because I know how to get a great result by using this, these AI tools. I know about the AI tools. I know I know how to direct them, and so I think that somebody with that skill set will be a monster if you have great judgment as well. In other words, the ability to just assess risk and know, hey, let's proceed with this. Or, you know, let's go that direction that'll never change. Folks who do that and who have very good judgment, capabilities and skill set are will always be winners, in my opinion. Combine that with the ability to direct AI the right tools, then it's over for everybody else.
Sam Pepper 48:19
So I always ask everyone the same question at the end of this podcast, which are, what are your three favorite buildings in Los Angeles, I'm going to actually ask you two questions. So that's one of them, what are your three favorite buildings or places? It can be place. The other one is, what are your three favorite AI tools right now?
Jake Heller 48:37
Three favorite buildings? Well, the obvious one is the closest in and out to where I live.
Jake Heller 48:44
Surprisingly, no one has said in and out before.
Jake Heller 48:44
And then, I mean, look, the buildings that we've developed, that my family's developed, they're very they're very close to home, special, right, blood, sweat and tears. It took a lot to get those up and built. So there's too many account. But if you want to be specific, we built a 40 unit building over in the Beverly Hills area that we wrapped up, I don't know, two or three years ago. Great location, great asset, great building. We'll be in our family for generations to come. So I think, I think that's pretty special. Moving over to the AI tools, my favorite at the moment, Claude. I previously was a ChatGPT junkie. It was always open on my screen, and in the last week, since claude's released, Sonnet at 4.5 I've been all about Claude.
Sam Pepper 49:31
Why is Claude better than ChatGPT?
Jake Heller 49:34
I don't know if it's better with these tools. It's almost like you got to feel it, you know, like these are almost like human beings in a way, like, Who do you vibe with the most? Right? And so Claude, like the way it responds, the way it understands me. It's research capabilities, the way that it executes code and creates files like it just does it in a way that resonates with me more than chat. GBT, at the moment.
Sam Pepper 49:59
You're sounding a little like Joaquin Phoenix's character in Her a little bit there where it's like, I'm starting to vibe with this AI agent quite a bit.
Jake Heller 50:07
It clouds my AI girlfriend at the moment. But it's funny, because these tools, like they always come out with new updates. It's every week, and so it changes. I've bounced around between these so many times, but for me, it's interesting. I subscribe to all five large language models. There's a lot more, but the five main ones, chat, GBT, Claude, Gemini, grok and perplexity. It's expensive. I wouldn't recommend it to everyone to subscribe to all five, but I know what each one of those serves my purposes for which one I prefer to use, and why.
Sam Pepper 50:36
And so Claude is one, and then what are the other two?
Jake Heller 50:38
I think Manus. I think Manus is awesome. It's an autonomous agent, so it could execute tasks and do quite a bit. And then gamma, it's great for declaration. And then, at the moment, endex and shortcut the two Excel super agents. I haven't decided which one of those I like better yet. We'll put that into one.
Sam Pepper 51:01
Jake, I really appreciate you jumping on. I think this is so informative for folks. You're at the cutting edge, the jagged edge of AI and real estate, and I think your platform is going to have a lot of success over the next few months and years. So thanks for jumping on.
Jake Heller 51:19
Yeah. Had a blast. Thanks for having me, Sam.
Did you love this episode? Let us know by rating and reviewing our show on Apple Podcasts. It’s easy - simply click this link, click on Listen on Apple Podcasts, scroll to the bottom of the page, and select “Write a Review”. Let us know what you liked best about the episode, and what others can expect when they listen to our show.
While you’re at it, consider subscribing to Building LA. When you subscribe, you can guarantee you never miss a conversation with one our renowned designers, architects, and developers. Subscribe now!