.png)
KP Unpacked
KP Unpacked explores the biggest ideas in AEC, AI, and innovation, unpacking the trends, technology, discussions, and strategies shaping the built environment and beyond.
KP Unpacked
KP Reveals His Next Big Ambition
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
What happens when a VC flips to the founder side and raises $13.8M to fix the biggest broken relationship in construction?
In this episode of KP Unpacked, KP and Nick finally reveal what's been hiding in plain sight across 20+ podcast episodes: Zero RFI, KP's human-first AI-scaffolded platform company purpose-built to modernize the construction industry at scale. After spending two years in conversation with General Catalyst – not shopping decks, just iterating on conviction – KP was handed $13.8M and a mandate to solve the asymmetry of information that leaves owners helpless: architects billing by the hour, contractors burying change orders in in 400-notifications floods, and buildings delivered 80% over budget.
The reveal unpacks everything: why Zero RFI and why now? why Zero RFI isn't SaaS (it's people backed by AI toolboxes), why scaling means buying 50-person firms rather than chasing enterprise sales, and why the owner's rep model is the only position with enough leverage to actually drive industry change. KP breaks down why BIM failed owners 15 years ago, why most construction projects run 80% over budget (McKinsey data, not hyperbole), and why his biggest technical risk is Anthropic releasing features that render what his team just built obsolete. The through-line? Technology has created deflation in virtually every other industry – construction remains the exception, and Zero RFI might finally be the answer.
Key questions answered:
- How did KP raise $13.8M from General Catalyst without shopping pitch decks?
- What does "human-first AI scaffolding" actually mean for an owner's rep?
- Are most construction projects really 80% over budget?
- Why do owners suffer from information asymmetry against their own vendors?
- How does Zero RFI scale—buying companies or SaaS sales?
- How does Zero RFI become distribution for Shadow Ventures portfolio companies?
- Can you actually break the billable hour model in AEC?
If you're an owner tired of 80% budget overruns and zero accountability, a VC wondering what happens when your partner becomes a founder, or a startup trying to crack owner distribution, this episode reveals the playbook for leveraging AI scaffolding to fix construction's most broken relationship.
Listen now.
BuildingWorks & Brookwood Sponsors
Hey Nick, how's it going?
SPEAKER_01KP, good to see you again. Covered fully from the flu. You and a lot of people, man. Down for the count. I'm glad that I earned a table at your dinner party conversation on whether or not I'm I'm healthy.
SPEAKER_00Well, I mean, that's the whole thing. Like we're all doing, I don't know. I wouldn't say that we're biohacking, but we all do various things. Modern things to do. Yeah, you know, we're we're fasting, we're eating a lot of proteins, like all the things, you know, nicotine, like all the stuff that people are like doing. Um, and it creates some divisive attitudes towards like what is healthy.
SPEAKER_01Yeah. I think the most interesting part of that discussion of what is healthy is like agreeing on a standard benchmark for the right things to measure. Like you look at Bryant Johnson from Don't Die, you look at, you know, your standard medical doctor, or even, you know, in you get into Eastern medicine practices and like everyone's definition of what is what is healthy is like pretty drastically different. So that's yeah, it's always a a fun point of debate. I'm like, what do you what are you optimizing for? And like, yeah, how do you like, you know, some some people, for instance, they they might like longevity with Brian Johnson, they might not care about optimizing to be 120. They might just want to optimize for health in their, you know, in their in their in their early years, right? And so how do you yeah, how do you how do you square that when like the time horizon for you know desirable age is is drastically different?
SPEAKER_00Yeah, I'm generally kind of a moderation person, right? And don't don't get too stuck on any one thing. But but also like, I mean, if you enjoy being in the gym for two hours a day, by all means like being the gym for right, yeah. But if you add up that two hours a day on your lifespan and how much time you spent in the gym, there's a good chance like while you've extended your life by so many years, you probably are not winning, right? You it probably makes up for all the time you spent in the gym.
SPEAKER_02Yeah.
SPEAKER_01It's like the debate on what is even good exercise, lifting, running, doing playing pickleball. You know, everyone has a different, you know, exercise regime of of interest. And you get oftentimes people get people get shamed if it's not the right, most healthy one. And I think that's like the most ridiculous thing in the world. Like exercise is should, you know, should should be should be fun and something you actually want and look forward to, I think.
SPEAKER_00By the way, all this stuff should be highly personalized, right? Just like how medicine is moving towards high personalization. If you look at like the rare disease things that are happening now because of AI, you're like, okay, we can lower the cost of coming up with a cure for a uh for a you know, a rare disease that doesn't hit a large population. It's all becoming highly personalized. So I think about health. Like, I mean, we are all so different. Mindset, biology, you know. I have bad knees from lots of soccer and lots of tennis. I shouldn't be running. Like, no one's gonna convince me to be run to run, right? So, no, I think that's you know, it's it's it's all a personal choice, and it's all like, but that's why you that's why the debates arise, right? Because everybody thinks that their idea is the right idea.
The Zero RFI Startup Reveal
SPEAKER_01100%, yeah. Well, we got a fun episode in store today. A reveal, a reveal of sorts. A reveal, the thing one might call it a reveal, but yeah, I think by the time this episode airs, you will have officially launched your new your new company, your new startup. That's right, which is called Zero RFI.
SPEAKER_00That's right. So give us been tough keeping it under wraps, man. Like as much as I publish and we pod and all that, like I think, you know, pat on the back, we've we've been able to keep our mouths shut pretty well.
Raising A Seed As A Founder
SPEAKER_01It's pretty funny. You could probably listen to the last 10 to 20 episodes and immediately map out exactly what what the strategy is. And I I actually think I'm curious I'm curious to get this feedback from listeners, but my guess is that there are some listeners who are gonna have a lot of aha moments, yeah, who yeah, can you know immediately understand like, yeah, why you've chosen to do what you're you're doing and also be like, oh, this is what he's been talking about for the last six months and what he's been like indirectly alluding to. But yeah, well, first of all, congratulations. I mean, it's I know it's been a Herculean effort to get it to this point. The announcement next week is the launch is also a fundraising announcement. And so how much money did you raise?
SPEAKER_0013.8. The PR people say I can round up to 14, but um let's stick with 13.8 for accuracy.
SPEAKER_01Okay, 13.8, good, yeah, solid. Yeah, not bad for a seed round, huh? What's it what was it like being on the other side of the table raising money as a founder again versus being on the investor side?
SPEAKER_00You know, I think it's it's super interesting in that you know, it's not like I went and shopped a deal and shopped a deck. This was I would say, like when we talk about our firm growing up and maturing, right? We're in that weird teenage years, I think. I would definitely like to operate the way my partnership with General Catalyst kind of evolved, right? It it wasn't like, oh, let me put a pitch deck together and pitch a bunch of ECs. It was a series of conversations with Paul Kwan and Haymont and Mark Bargova to just say, like, is there a there there? Is this, you know, is this industry ready for change? And, you know, posing that wild question, right? If you had access to unlimited capital, what would you do? Which I don't know that we've ever asked any of our founders that question. You've uh you've asked that question. I probably asked that, I do ask that question, but it also feels you know, when you're not general catalyst, it feels a little disingenuous, right? Like, why do you ask KP? Do you have access to infinite capital?
SPEAKER_01Definitely, yeah, it definitely kick sometimes it kick kicks the bruised knee of the founder who's struggling to raise money when you give them a few. I'm trying to raise 500K, man. 100% unlikely to occur.
SPEAKER_02Yeah, yeah.
SPEAKER_00I'm trying to make payroll, leave me alone.
unknownYeah.
SPEAKER_00So I think the process, I I would love to say, like, oh, I, you know, it was a classic founder process. I I think it's it is a process that I would say, I don't I don't know that it was unique to me. I think in the world of second or third time founders, I think it's it's a fairly common thing. Like, hey, you've been successful, you know what we're doing. What do you want to do next? Right. I think that I think that's that was kind of like a similar conversation that a you know a a serial founder would get, right? So I think it wasn't that different. I think it was super different than like the conversations we have at Shadow Ventures with founders that we're talking to, right? So, no, I think, you know, when you look at that, I think it's what's interesting about that collaboration, and I it's the first time I've experienced it, right? So it's I'm not saying like, hey, this is how it works. I'm saying this was my this was my experience, but I understand it's similar to other experiences. I think when you haven't decided what you're doing, but you're very you have strong conviction about the problem sets that you want to work on to be able to have early conversations with the VC to shape and mold the idea, I think is really the ideal way of doing things. Because I think otherwise, you know, someone pitches us for an idea, it's a binary outcome, right? Either we're interested or we're not, either yes or no. So, you know, this was like a couple year process, right? Of relationship building and conversations and you know, sharing thoughts back and forth and you know, seeing companies get funded and asking the question, like, well, you know, why did they get funded by so-and-so? Like, why did that make sense? Why did it not make sense? Like, what blind spots are they missing? But I think, you know, it's a definitely a different process. But I will say this I wish founders would come to us sooner, right? Like before you quit your job, come start talking to us. Right. And it might it might be that you're at your job for a year while you're talking to us, but I think the outcomes are probably better because we understand kind of what the capital markets look like and availability of capital and trends and all that. And the founder can iterate the idea because I think this idea of for our industry specifically, and there's other industries, you know, let's go build an MVP and see what happens. I just don't, I'm you know, I've said it before, I don't think MVPs work in our industry very well. So, my experience, I wish more people could experience it, right? Which just means you have to have conversations earlier. And those conversations may go nowhere, right? It might be some someone calls us, hey, I'm thinking about leaving my company. Can we start chatting? You may end up like, hey, there's nothing to do, or you're not the right person to do it, or you know, whatever it is.
SPEAKER_01Yeah. I I'll I'll add to that. I think that you know, for for an in uh for for an investment firm, for a VC, that has to be a mindset that you that you also take. Because I'll I'll admit, like, you know, we there's so much that we see, right? There's so much inbound and you know, people reaching out about ideas and LinkedIn messages and introductions from other investors. And so you get you get a lot of you get a lot of noise. And some of them are like are good matches for our industry and are and are compelling, and others are just like not a good fit, and you're just like sifting through, you know, the paperwork of telling people, hey, you know, this is what we do, not a good fit, blah, blah, blah. And you kind of get worn down by that process. And so there's sometimes like, I'll get, I'll get an inbound from someone and they'll be at that stage. And admittedly, I'm not as open as I should be to having that conversation of like, hey, you're at so-and-so reputable AEC firm or so-and-so reputable tech company, and you want to start a construction tech focused startup. I have to remind myself, like, those are the types of conversations we want to have, even though it's super early and there's no immediate action that's going to be taken. And I think for any VC that a founder, a prospective founder talks to and is trying to work with, like I would, I would vet for that. Like, try try to see if they're open to doing that and open to basically brainstorming the idea with you a bit and yeah, spar sparring because I think like you get you do get a lot of there's both the benefit, the relational benefit of getting involved early and get some getting someone, someone's opinion you respect. And then also we get to see you move, right? We get to see how much progress you make in 30 days and 90 days. And that tells us a lot in terms of, you know, when you compare that that cycle to a uh two-week diligence cycle when you're running a fundraise process, like it's night and day the difference in terms of comfort that that the VC can have.
SPEAKER_00No, I think we also like one of the red flags we saw, and listen if we talk about the use case of a founder leave leaving a corporate to go to a startup, you know, one of the flags we always have is like, hey, they've been working at a corporate with PTO and that like, are they gonna be able to work in a startup situation? And you know, if you can hold down your 40-hour week job and work on your startup 60 hours a week, you're starting to show us, right? You're starting to show us that you have the grit and requirement to like once you quit your day job, you'll you'll come on board and and and run like a startup versus you know running like a corporate person. But I but I think you know part of the challenge is you know, catching people where they have conviction about the what, but not about the how. And I think too many people show up to us where they already have, you know, I remember when we first started this process, Hammond was, you know, we're we're similar ages. I think I'm a little bit older than him. But you know, he kind of coached me on like, hey, you you know a lot. I'm not gonna take that away from you. But the thing you need to watch out for is acting like you know nothing, right? Just keeping an open mind to not knowing everything. And and and we iterated, right? We iterated several times on our thinking. So I think I did a pretty decent job of not being like, I've been here forever doing this and blah, blah, blah. You know, that but I think we have some founders that come to us like I was at XYZ construction for 20 years, I know everything. And, you know, call it the beginner's mindset or whatever, you know, whatever the latest book is telling us about these ideas. I do think that was some great advice that Hammond gave me to say, like, you know, I recognize you know your stuff, right? I'm not gonna take that away from you. Also, don't get stuck, right? Don't get stuck.
SPEAKER_01That's really good advice.
SPEAKER_00Yeah.
Owner Ally Across The Lifecycle
SPEAKER_01Yeah. Let's talk about what zero is. We haven't covered what you're actually doing yet. So yeah, walk us through the walk us through the idea. There's obviously a lot to lot to dig into on the business model, but yeah, what is zero RFI?
SPEAKER_00Yeah, so we are a human-first AI scaffold owner's representative. So the idea is we will go to owners and help them get their buildings scouted, designed, built, and operated and managed, right? So full life cycle from beginning to end. So if you're thinking about building a building, you hire us and we run you through the full full life cycle, which means we have people, it's not go to this website. However, our general belief is that our our team has a set of tools in their toolbox that they get to pull from to do this analysis to move quickly, right? To move quickly. And I think part of it is we have not seen anyone really be the owner's ally for the full life cycle, right? In other words, worrying about operating costs before you design it, like thinking about all these things. Sophisticated owners tend to do it themselves, but you know, if you're not building a building every day, like how you know, there's it's very difficult for an owner. And so I think for me, you know, people, you know, when I look at like how did we end up here, right? It's a little bit of like all the ideas that have come before this that I finally felt like where I saw AI trending was that AI being less of this enterprise bucket that does everything, and really the ability to have to create facile tools for every situation to like create transparency. And quite honestly, like there's such an asymmetry of information between owners and their designers and contractors and product manufacturers to kind of level the playing field for the owner. And I don't think anybody's focused on that. I don't think anybody's focused on that at all. And that's where when we talk about all this construction productivity data and all that, you know, the problem is the only people really suffering from this lack of productivity are these owners, developers, and then society writ large. I don't think the industry is actually suffering from it. If you ask an architect, has this lack of productivity affected you? They're benefiting from it, actually. Build more hours, right? If you ask a contractor, are you benefiting from this lack of productivity? They're doing fine, right? They're getting paid. So I think what it takes, and because of the fragmentation of the industry, you have to have an entity substantial enough to drive this behavior. And it's very hard for an owner that doesn't do these things every day and doesn't have the tool sets. I mean, go out there and look at AI tools. There's there's a hundred tools for how to do pre-construction. There's a hundred tools, AI for scheduling. If you think about the weaponization of AI, right? How's the owner going to deal with this? Like all those tools are not driven. The goal of a pre-construction AI tool is not to give the owner a better building or a cheaper building or a faster building, it is to enrich and reduce risk for the contractor. Theoretically, I think we talked about on a previous podcast about kind of a little bit of a prisoner's dilemma. If they actually could reduce costs by 30%, is a GC gonna really share that with the owner? Or is it more to your point, like, well, who's ever flinches first, right? Whoever that contractor is that bids 30% below everyone, that could drive the behavior. But we haven't seen that, right? And I think, you know, I wrote a book 15 years ago now, or longer, I can't even remember, 2011, so 15 years ago, that was called BIM for owners and developers. And if you read that book, my answer then was if owners and developers own the BIM process and orchestrate this, they can win. They can have a cheaper, better, faster building, and they kind of win. Instead, what happened that never happened, right? The architects used it, the contractors used it. And generally speaking, the owners have seen zero benefit or very little, maybe in facilities management, a tiny, but developers have seen no benefit from BIM. If you ask a developer about BIM, they're like, Yeah, my architect uses it. What do you want to know about it? Right. But my view of BIM when it first came out was oh my gosh, owners can now have a level playing field of information because they don't have to know how to read drawings, right? They can see the 3D, they see what they're getting. And I used to say this is kicking back from years ago, right? When I looked at BIM, the largest purchase ever made, unseen, with no performance criteria, with no understanding of total cost of ownership, is a building. In our own consumer mindset, if you came to my car dealership and said, I want to buy a car, and I started sketching some stuff on napkins and said, Here, Nick, buy this car. It's only$100,000. You'd be like, Whoa, what color did like you'd ask all the questions, right? Like, no, no, sign here, it's$100,000. And oh, by the way, like, I can't tell you how many miles per gallon it takes. I can't tell you what the maintenance is. And then when I deliver it, you're unhappy with it. And it was also instead of a hundred thousand, it was two hundred thousand. There was a stat that my PR team shared with me. I was like, oh, that can't be right. I was like, that sounds aggressive, right? And it was a McKinsey report or a Deloitte report, you know. But they said that most construction projects are over budget by 80%. And I heard I saw that and I was like, oh, that must be a typo. That's gotta be a typo. That sounds right to me. I I thought it was aggressive. I thought it was like 80%. That's like I mean, if tomorrow your bills went up 80%, how are you living that life? Like, how do you plan? It's a problem, yeah. Yeah, right? Like it's a problem. We can't even, I mean, look, we can't deal with eggs being 10 bucks a dozen, you know? Like, so I'm thinking there going, and it was this was just yesterday. It was fascinating because I was like, oh, it can't be 80%. I know it's bad, guys, but are you sure you're not being a little bit you know, aggressive to get the press's attention? They're like, nope. And they sent me a screenshot of where it was cited from. I was like, wow, it is that bad, right? So I learned something new yesterday. It's it's I thought it was bad, it's like horrific, right? So but I think that's ultimately what we're trying to do is if every other industry, and to your point, I've been throwing around little breadcrumbs, right? Why is it that every other industry, the technology has had a deflationary effect in that product or service, but not in construction? And so, you know, my general thinking is like we might be that we might have that answer. Can how can we not how can we have better predictability, right? So, you know, my view of the world is risk management as a human job, uncertainty is a great job for AI.
SPEAKER_01So if I'm if I'm understanding correctly, the decision to be an owner's rep was primarily to deliver to deliver either a a lower cost option to the owner and or better services for the first time, knowing that if you built a product that still made strides to improve the overall construction process and the overall AEC process, getting getting something built and developed, but you're selling. Into the individual vendors, it doesn't allow you to actually impact in in a lot of ways, maybe indirectly, but it doesn't allow you to directly impact the the owner experience who is the customer.
unknownRight.
SPEAKER_01Is that is that the right way to think about it?
Human First Team With AI Tools
SPEAKER_00Yeah, I think the right way to think about it is a couple of things. Like, you know, owners don't build often enough that they're going to get up to speed on a set of tools. Right. So it's it, there's never really a good option for a self-service model. I think the other thing is the experience is horrible. I mean, I wrote an article about this about how do we bring joy back to this profession from an owner's perspective. Right. I mean, if you're a school principal and you just got approval to build a brand new middle school, you should be excited. They're excited at the very first meeting. And then after that, it's nothing but bad news. And it's bad news delivered very cold, right? It's like, well, that's what things cost. Yep. Sorry, you can't have that extra classroom because it's not going to work. Oh, you can't do it. It's basically a bunch of like, you know, no, you can't do that. And it's like horrible, right? It should be very different. So I think there is also like the customer experience in this industry has not been has not changed or been reshaped ever. So we think it's not just about predictability of costs and schedules and getting what you want, uh, you know, getting what you paid for. It's also like, why can't it be, you know, fun along the way? Right. It's, you know, someone I was talking to someone and they said, you know, there's not a kid out there that doesn't love Legos. Almost every kid out there loves Legos. And if you've never built a building, you kind of like it's it sounds like a great idea until you do it. And I've told my wife, it's like, we will never build a custom house because we will probably not end up together at the end of it. I'll be a custom house for someone else. Right? It's not fun. It's it's it's painful, it's brutal. So I think that's another part of how we're looking at it is like, why why can't it be fun and interesting and collaborative and all that?
SPEAKER_01I think you said human first with AI scaffolding was how you describe zero. Talk me through your definition for that. What does that mean?
SPEAKER_00So if you're building a hospital, I'm gonna deploy people in your in your office, like to work with you. And and really that the that team is backed by a set of AI tools that they can pull out of the toolbox as needed when needed. And so we're not trying to take away, you know, it's kind of interesting, you know, once we fully launch by this time, you'll probably see a hundred job openings. So we're not this like, let's get rid of people and AI does all the work. What we're saying is let's enable people with deep tools that remove the asymmetry of information and puts it more in their position, and they're just better equipped to handle things. Look, if you get a hundred, if you're a project manager on a project and you get a hundred pro core notifications and 50 emails and all the reality is you cannot, you know, you cannot take that information and make sense of it. There's just no way. And that's where a lot of this asymmetry of information started was a lot of the technology that allowed the vendor system to create lots of content quickly, right? In other words, here's a set of drawings, here's a set of PDFs, here's a BIM model, take a look, review it, get back to me tomorrow. We need to make decisions. In other words, I'm gonna wait till the last minute to send you a deliverable, and that deliverable is gonna be 100 sheets. I'm gonna wait till the last minute, like, hey, you need to approve that change order or else the project stalls. Well, I haven't looked at it because it's buried in 400 notifications, right? So there hasn't been a good way to ingest, because I would argue a lot of that, you know, like we were talking about a simple idea. Every project on when it's in construction, there's a daily report that is published. It actually has a lot of great information. And if it doesn't have a lot of great information, you can ask for that information, right? You think owners are reading those daily reports? You think the current owners' reps that are dealing with all this paperwork are looking at those like, no, they're not. So that's where from day one, if you look at the compounding effect of asymmetry of information, it just compounds along the process. Right? We go from three watercolor pictures of what the project might be to stacks of submittals to just all kinds of, you know, so it's the beginning of the end, right? It doesn't get better through the process. It just gets worse.
Breaking Hourly Billing Incentives
SPEAKER_02Yeah, yeah.
SPEAKER_01We've talked about how in the future AI has the potential to break the billable hour for specific vendors that are operating, you know, with a billable hour infrastructure, like engineering, like architecture in many cases, like the owners rep model. Do you think that like do you already see that happening? Like when you're building product and interacting with, you know, your your first acquisition, which we should cover in a second. But the like do you see that that infrastructure is breaking down, or do you see opportunities to break that entirely, or do you think that remains intact for a bit and it happens more gradually?
SPEAKER_00Well, I think it starts to bring, I mean, hopefully, I mean, look, uh, I as much as I'd like to be the winner takes all in this business, right? Hopefully other people like see what I'm doing and copy me. It would be a good thing for the industry, right? I think one of the biggest challenges when we talk to engineering firms, architects about billing by the hour. I think some of them disingenuously say, like, oh, I wish I didn't have to bill by the hour. I think they actually like it, right? Because when you bill by the hour, there's a lot less accountability to the deliverables. You just bill hours, right? But I think there is a the much larger segment does want to get away with it. When you ask them why can't they do that? They say because their end customer will not let them do that. And so if you look at the whole RFP system and how you know rate sheets and all that, that is because owners kind of don't have a choice. They don't know how to evaluate, right? They have to make everything apples to apples to apples, and it's a procurement process. Our hope is there's this idea called you know qualifying qualification-based selection, QBS. It's a joke because at the end of the day, all the engineers and architects know it's really just gonna be price, right? So if you think about that, us as an enlightened owner, if you're a GC and you show up to me and you say, like, hey, here's what I want to do. I'm gonna 3D print this, I'm gonna modular that, I'm gonna have robots over here. I'm gonna be like, tell me more, right? My team is gonna be like, tell me more, help me understand this, right? A traditional owner or owner's rep is like, I don't even understand the words you're saying. Please just give us a bid. Right. And so I think a lot of the, you know, a lot of people take my words as like I'm trying to beat up the industry. What I'm saying is I'm gonna clear the road to do things better. I actually was gonna work.
SPEAKER_01Yeah, I was gonna I was gonna ask you about this. So your intention primarily, first and foremost, is serving the owner and and delivering a better experience overall for them on time, on budget, getting getting closer to the experience they want. But yeah, like the the other side of that is how do you better serve the AEC industry and the real estate industry? And so yeah, like do you think do you think about that from a from a product standpoint? Like is that a product driven exercise, or is that is that a is that a is that an organizational change that you would make to you know to support the industry better? Like how do you think about that?
SPEAKER_00Yeah, I think there's a couple things is one, you know, if you're a carpenter and you're gonna come do framing and you show up with a hammer, like, sorry, move along, sir. I need someone with a nail gun, right? Like there's just certain things. So I think the owners have actually many ways, because they they have to they're trying to reduce the variables of risk, they tend to default to like, here's the way we've always done it, right? Which is fair, right? When you're when you're when you when you don't have a lot of choices, that's what you come down to, right? So it ends up being apples to apples, what's the price, right? So I think my approach is we actually have a program we'll be announcing that we're launching that really allows people that want to serve our owners get the opportunity to interface with us in a different way, right? In other words, if you are more tech forward, there will be a benefit to working with us because we can tell, right? We, you know, it's kind of like if you talk to most owners' reps right now, it's like, hey, what are you doing with the BIM model that gets submitted? Yeah, we look at it like we went nothing, right? They do nothing with it. Whereas, you know, we might be like, hey, if you're gonna submit a BIM, here's how we want you to submit it. And we and our AI will actually check it. So you better deliver something good, right? Don't don't put garbage in there. So I think my view is you know, the example I use is if you look at what Walmart did to their for their supply chain, everybody thinks about Walmart and said, well, they got the best prices out of all these people, which is true, but a lot of the way they got their best prices is by giving them a vision and you know insights into their own business that they didn't have before. They didn't know, right? I don't think there's any drywall contractor out there that wants to throw 30% of the drywall into the dumpster. That doesn't help them at all. They don't, you know, that's not helpful to them. They don't want it. I mean, the good ones, they don't want that. So it might in fact be that through our program that, hey, here's some tool sets you ought to think about implementing that will actually reduce your waste. And if you can go do that, we'd love to give you a shot at the next jump, right? So I do think that, you know, while I'm kind of saying like the owners, the only owners are first, and they are who we all work for. But the owner's ability to give feedback to their ecosystem, to their vendor network about what good looks like and what is better to start to look like, they can't. It's just not realistic, right? So I think even there, you know, I had an engineer tell me that's already joined our program. He said, look, we deliver when we do structural engineering, we deliver it three ways: reinforced concrete, steel, and wood, because we want the client to have optionality. Our competitor does not. And when the owner is looking at selecting structural engineers, we get no points for providing such optionality. They literally come back to us and say, well, it's gonna be, you know, everyone else like bid on it for reinforced concrete. Just change your price to a denominator.
SPEAKER_02Yeah, yeah, yeah.
unknownYeah.
SPEAKER_00So let's take, let's, you know, let's take creativity, ingenuity, and drive it to an RFP, which is basically what? Mediocrity. So I think in many ways, we are our our the change rate just won't be in like governance, oversight, data, and all that. It's that feedback loop of like, hey guys, have you thought about this? Our data shows that do to do to do here's a tool you might want to go use from a third party. Like, I mean, we already have one deal right now where we recommended one of our portfolio companies at Shadow, like, hey, go work with them because that'll help you get the data we need back better, cheaper, faster. And by the way, it's probably gonna help you execute the work. And they were like, I've never heard of this. You know, I've never heard of this. It's like, great, good opportunity, right? And they got very excited. And I think they got very excited too because they because we're looking long view of like the lifecycle management. We're like, yeah, and by the way, that data is gonna help us maintain your building better. And they're like, oh my God, you're right. That's so good. Let's do it, right? So I do think the the the change rate within the ecosystem of AEC, I think it's gonna be very all this money that's been put into innovation in these firms, and they haven't been able to get much differentiation. I think we're gonna start moving that along.
SPEAKER_01To bring it back to the to the name you were just alluding to, I think the ultimate goal, which is a zero RFI state for the industry, which I think, you know, in theory, that gives you a way more predictable schedule. It gives you a lot more predictable cost structure. But yeah, like what does that what does zero RFI mean to you? And maybe like maybe talk me through the the time frames of which you're thinking, like that's actually achievable.
SPEAKER_00Yeah. So I think like it comes back to what I built several years ago back in 2005, which was we did constructibility analysis doing using BIM. And the whole idea is if we used BIM to try to build the building before the contractor was hired, we could find all what we called then was virtual RFIs. And we put them all in a discrepancy report. So the idea was before you put it out to bid, go clean up all these constructability issues, and you're more likely to get better bids. Preventative. It's preventative, right? So let's simulate construction, not just review the drawings for detail issues and misspellings and all that, right? But actually, based on this these plans and specs, can I build this right now? The problem with that back then was it was very manual, right? The other thing was there is a body of knowledge in construction that you don't have to be as prescriptive, right? I don't have to tell you how to swing a hammer, I don't have to tell you where to put the nails, right? That's kind of your job. And what I found is with AI, the ability for AI to determine like what is not built constructible against an inference model that says, well, the contractor knows how to use a drill, right? That's commonplace. Whereas before it was like, I mean, we had to act like the contractor knew nothing, which means we had 30,000 RFIs. And then in practical, you know, half of those were like, I mean, if you're a contractor, you know what to do here. It doesn't have to be as explicit, right? So we always used to say errors are easy, omissions are harder because they don't exist to be corrected, they have to be inferred. And when you go down this trap of like inference and inferring things, it means you have to have a certain level of knowledge to be able to infer them. And back then when I was doing it, we didn't have that. You know, we had a bunch of experts sitting around, like, I mean, would a drywall contractor be able to understand this? I don't know, man. What do you think? And it was like debates, right? And you'd have someone say, like, I mean, if they're any good, they should know what to do.
unknownRight.
SPEAKER_00And it's like, well, that's not how to like, what does that even mean? Right. So I think the mission of zero RFI is that everywhere along the process, you're operating without questions, right? I don't have to ask you for information. It is there, it persists, it's available, it's clear.
SPEAKER_01Do the back to your point on the on the drywall contractor are are are the are the models good enough today to infer correctly on the knowledge of the of the of the average sub that you're working with?
SPEAKER_00Not alone. I mean, that's why we we've spent, I mean, we've been in stealth mode for over a year now, right? We have we haven't been in market or we've just been building heads-down building stuff, right? And so and I will say that uh versus a year ago to today, the LLMs out of the box are doing some things that we you know we spent a lot of time building that now it just does, right? Which is fine. That I mean, I always say, you know, when I talk to Barry, our CTO, I was like, our biggest risk of technical debt is that whatever Claude releases, whatever anthropic releases tomorrow that we just got done building, you know, gets uh, but that's that's why it's you know, that's why we're venture-backed, you know, these things will happen. But I know I think it out of the box, I don't think you'll ever see it per se, because I think it's this is where understanding how to take, you know, this this is a true thing, although overstated a lot of times in our industry, every project is different and unique, right? So when you look at that, can an LLM determine if you're on your fifth identical Waffle House, can the LLM decide, understand what's you know, what's happened, right? What's happening in terms of, well, it looks like here's your fifth Waffle House and I found all these problems, right? Because I learned off the first four. I do believe that a lot of what we've built is really driving how do you think, how do you look at every project being unique and creating inferences? Not two no two airports are not alike. If you look at a new construction versus an addition versus renovation, there's lots of unforeseen conditions, right? But some of those finding those unforeseen conditions are kind of known. Like here's what you should be looking for, right? I don't see the, I mean, look, if we could just operate and just pay anthropic$300 a month per head, I'd be thrilled, right? There's nothing wrong with that. I think we're far, far away from that. And and maybe never, right? Because when you think about every project, every building being a new idea, how much of the existing corpus of data is really going to help you inform how you build that next new thing.
SPEAKER_01Everything's a markdown file, man. What's that? Everything's a markdown file now. Yeah.
SPEAKER_00Well, I think it's like, you know, in the legal profession, right? One of the things that's, and this is where things are different, right? In the legal profession, there's a thing called case law. It's open sourced. It's published, right? So if AI can read all the case law and start to determine like what is, what has precedence, what doesn't, even here's how you should argue, right? To win, that's probably fine. It's probably gonna get get there very quickly in the next six months, even, right? But what when there's new a new case without press without enough direct precedence? That's a big question everyone has.
Scaling Through Acquisitions Not SaaS
SPEAKER_01Just no markdown file or skill for that. Right. Yeah. How do you scale this thing? So owners rep, so human first owners rep with some AI embedded. Yeah, what's the processor scale? Buying companies.
SPEAKER_00So we're part of uh what's that?
SPEAKER_01Sounds pretty straightforward.
SPEAKER_00Yeah, pretty straightforward. We're part of uh GC's AI store AI roll-up strategy, right? And so our go-to-market is buy companies. But I think unlike PE, we don't have a oh, we have to flip this company in five years or else. We're not we're not operating with any of that stuff. We're basically like Evergreen, right? So there is no we have to sell exit by a certain date. That's not that's not the priority. We also think so. I think we're competing against private equity that's trying to buy these firms. I think if someone wants to sell their firm and retire, we're probably not the right partner. If you're a firm that sees where things are going and realizes I don't have the technical resources to do this myself, I don't have the capital resources to triple my business in three years. So we become an ideal partner in those situations. But I think if you look at there's a strange thing, and we studied it. I mean, we've definitely been studying a lot. It turns out that owners, reps, once they get over 50 people, the quality tends to go down. The capabilities tend to go down. So there's a lot of great 50-person firms that are very passionate about delivering excellence to their customers, and in fact, have not grown over 50 by design because they're like, we just there aren't enough good people to do this. Our point of view is I don't need everyone to be expert level. They're getting a lot of help. You know, so when you look at some of these big commercial real estate firms, the the sad thing that they say is, KP, you're so fixed on fixated on people being able to read drawings, half of our people can't read drawings. I'm like, well, that sounds horrible given that you're doing construction management. And they're like, well, it's it's not important, right? And I just cannot live with it. I don't understand what that what do you mean it's not important? Right? It's like, I'm gonna be a food critic, but I have no taste buds. Like, I don't, I don't understand, right? It doesn't compute anyway for me. So I think what the tooling we're building is I think, you know, maybe you're great at finance, maybe you're great at customer relationship management. You actually try to do what's best for your customers, maybe you're great at managing a meeting, but maybe reading plans is not your strong suit. That's what our AI does for you.
SPEAKER_01Let's talk for a minute about how this all re Relates back to shadow ventures and the benefit potentially to every current portfolio company and future portfolio company. Personally pretty pretty excited about the opportunity for a pretty big distribution advantage. So yeah, to just tell tell everyone how you're thinking about that.
SPEAKER_00Yeah, I think you know, most of our companies, right? They're seed, pre-series A, Series A, right? They're they're early. If you talk to most of them that are selling to contractors, to you know, et cetera, everybody's trying to chase like this enterprise agreement. And so if you're a startup and you go directly to the owner, one, the sales cycle is long. You know, getting a hold of the head of construction at Genentech or somewhere, it's hard. Takes a long road. Inevitably, they'll say, Oh, that's what you do, you should go talk to our contractor. Oh, that's what you do, you should go talk to our architect. So they get redirected down to the ecosystem because for these owners, once again, apples to apples, right? They're not they're not gonna bring technology to the table. They can only influence, hey, go talk to go talk to my GC. He might be interested. So we think we're gonna create a massive distribution opportunity for for our startups and other startups, right? That makes sense for the owner to kind of standardize on some things.
SPEAKER_02Right.
SPEAKER_00And I think that's what's kind of exciting because I've seen so many of our founders they they say things like, yeah, like if we could get to the owner, here's this new dashboard we have. If we could get to the owner, this is how this would benefit them. And they get zero traction to the owner. Because if you think about it, if you go hit one GC, that GC can bring you into a hundred projects in the next 12 months. In that same amount of effort and customer acquisition cost, you can go see one owner.
SPEAKER_01Well, at the same time, one project. They may or may not even care about using the tools, it's not their that's not their core competency.
SPEAKER_00Right. We've never used this before. We don't know. I mean, I'll introduce you to our contractors if they want to do it, sure.
SPEAKER_02Yeah.
SPEAKER_00So I think that's that's a little bit of breaking the mold. And you know, I've already talked to some of our founders that in the past had pitched me, like, hey, we want to go to the owner because we think we can add value. I'm like, dust that deck off. Let's talk about how you can add value to the owner.
SPEAKER_01Yeah, yeah, yeah.
SPEAKER_00Because in many ways, I you know, we are becoming the mega owner, right?
SPEAKER_01Yeah. So you're leaving the investing game, or are you you leave leaving us behind? What how are you? How are you pulling pulling off all the all the extra effort here?
SPEAKER_00You tell me how I'm doing.
SPEAKER_01You are still young at heart and more energetic than most people my age.
Competing With Rapid Model Improvements
SPEAKER_00So yeah, I don't think the good news is like, I mean, if I was going off and doing something in aerospace, totally different, right? I mean, I'm in the soup, right? I'm in the soup. And, you know, whether it be startups and I would say the only thing that I've become more biased by, and I think that's my own experience, it's just really pushing founders to think bigger. Like, I think what I'm not saying, I think what the relationship I have with GC is very special, unique. I'm not gonna take away anything from that. They're fantastic people. They I think at one point you were joking around, like, hey, you're gonna get like competitive term sheets. I'm like, no, man, like these are these guys are my partners. I'm not doing that, you know, like they're they're my guys. I think it has created a bias in me that like I want founders to think bigger. I think they can go have these conversations. Like I had, I mean, my resume as I, right? My resume as I, I think you can go have these conversations with these VCs and build bigger things. Yeah. So when I see things like, oh, we're a little featured, I'm like, come on, man. And and I'll tell you, where I've been really, you know, look, I think investing energy in AI, right? Using the tools and understanding how they work, it's like a whole nother job. I mean, you you know, I'm up at 4 a.m., 4 a.m. to 7. I'm either writing or coding, doing things, right? And so I find it fascinating when founders are not spending that time. You know, I've been, I mean, literally, I've been like, I see founder decks sometimes like this is not exciting. I dropped the deck in a clod. What would it take? What would you what would it take for you to build this product? It rarely says, oh, it'd be very hard. It's almost like, yeah, like this is pretty, you know, this is pretty easy, right?
SPEAKER_01Do you do you feel that you're insulated by the the the threat of model improvement and the threat that you're seeing right now play out in the public markets with any soft any software company? So most most of the SaaS businesses we've covered are trading, you know, trading, trading down quite significantly from where they were. Do you feel that the services dynamic protects you from that? Do you feel that you're better able to ride the the exponent of the models as a as a service provider?
SPEAKER_00Yeah, I think the services model is is a pretty is a pretty decent mode of sorts, right? Like we're actually doing the work. We're not competing, you know, you know, we're not competing against Procore and Autodesk, right? Like in many cases, they're they're already knocking on my door asking about how, you know, as they've things have have leaked here and there, like, hey, how can we be, how can we be a partner? How can we be a good partner? Because by the way, you look at Autodesk and Procore, just to pick on them for a minute, ask them how their owner initiatives have been going. Ask them how going to owners has expanded their business. And they'll tell you it's been hard. So it's it's it's hard enough for a startup. I mean, when Procore and Autodesk can't get the owners, you know, attention, and they say, no, go talk to my GC, they manage all our drawings, right? How does a startup even do it? But we we think that like change and influence starts there. So I think there's a couple of things. One, we're not SaaS, right? And I've I've done some demos for people, like, well, how much? I'm like, sorry, like that's not how we do, it's how this works, right? So I think we're fairly insulated from that. I think we're in a good spot from an acquisition perspective because the industry has started to understand this dynamic of being bought in private equity. And then I think our other advantage is that most private equity firms, maybe all, they operate on a three to your three to five year cycle, and then they have to sell. Right. So for me to convince Nick Durham at Nick Durham and Associates to sell their company that they've worked on for 15 years, that's very nuanced, right? It's very nuanced. But going to a private equity firm that already that already convinced Nick to sell five years ago, right? And me show like, hey guys, ready to junior disposition cycle. Yeah, and they go, Yeah, man, like let's talk.
SPEAKER_02Yeah, yeah.
SPEAKER_00So I think there's been enough seeds planted in the PE world that they're all at their three to five year cycle that we'll come along and pick up. Very cool, very cool.
SPEAKER_01Yeah. Chapter one to be continued.
SPEAKER_00Yeah. No, I think it's gonna be what I would say is it's been a lot of fun so far in terms of just working on these things. But I think it's also when you asked about how does this affect shadow ventures, I'm so deep in it, man. Like my ability to understand what's in market, what AI like I know you've been spending a lot of time in AI, but I think my I'm spending a lot more time.
SPEAKER_01You're actually building product, it forces a different, yeah, a different level of understanding and depth.
SPEAKER_00Yeah, agree. So I think you know, when we look at founders and they're like, yeah, we built this unique thing, and you know, well, what about what about the LLMs? Aren't they just gonna do this? Ah no, man, it's like, I don't know. Like, I don't know if you saw like the latest afternoon release from Anthropic. There's the morning release and the afternoon release, right? Like two releases a day, these guys are at, right? And it's like they don't, they're they're three weeks behind on like what's been going on. It's like, oh, sorry to tell you, like, Claude just launched a plugin that does exactly what your entire startup does. So I think we're in a good spot. I feel good about it.
Closing And The Anti AI Email
SPEAKER_01Exciting stuff. All right, man. To be continued, more to come on on Xero RFI on this on this pod, and and in addition to all the other KP ready editorial content you receive on a daily or weekly frequency. Up to 157,000 subscribers, man.
SPEAKER_00What's so funny is by the way, side note, I had a CEO of a construction company send me an email saying FAI. Fuck A fuck AI, right? And so I was like, sure, okay, right? He didn't unsubscribe though.
SPEAKER_01He was just mad, mad at the world?
SPEAKER_00He's just mad. But he still had the he was like Was he mad?
SPEAKER_01Was he mad that you were spamming him or is he mad about it?
SPEAKER_00He's just mad at AI. Gotcha. He was mad at AI. He's still a subscriber.
SPEAKER_01Uh send him send him a screenshot with the unsubscribe button. I was gonna send him I was gonna send him a gift bag with some candles and send him a subscription to Claude. Yeah, there you go.
SPEAKER_00Sounds good. All right, man. Thanks. Talk to you soon.