Artwork for podcast The Zweig Letter
Master Builders Reimagined: ElevateAEC Keynote - KP Reddy Explores AI, Integration, and Future Design Leadership
Episode 25626th February 2026 • The Zweig Letter • Zweig Group
00:00:00 00:31:23

Share Episode

Shownotes

We keep iterating on an old construct—let’s stop and start over. How do we want to do things differently? Let’s get back to first principles.”

  1. KP Reddy

Episode Summary:

In this special ElevateAEC 2025 edition of The Zweig Letter Podcast, we're sharing KP Reddy's keynote from the main stage—a straight-talk session about where innovation in AEC is actually heading. Drawing from real conversations with major owners, hands-on research, and his own experience as an engineer and investor, KP breaks down why the industry's current approach to innovation isn't cutting it—and what needs to change.

He tackles the questions that matter: What do owners actually want? How can AI and agentic systems make a real difference? And what would a new master builder mindset look like today? This keynote delivers practical insights and honest challenges for AEC leaders ready to rethink how they approach business, technology, and client value.

Key Takeaways:

  1. Owners Want Real Innovation, Not Lip Service: True innovation means solving problems in significantly better ways—not just adopting new technology for technology’s sake. Owners are craving AEC partners who listen and deliver true value.
  2. Communication Gaps Hurt Everyone: Clients frequently feel like outsiders in the process, with their feedback often ignored. Successful firms will prioritize transparency, owner-centric approaches, and collaborative requirement gathering.
  3. Tech Is a Tool—Not the Solution: Despite years of BIM and other advancements, core challenges with cost, schedule, and quality persist. The next leap forward will come from integrating AI, agentic design, and robotics as part of service delivery—and from business models, not software alone.
  4. Business Model Reinvention Is Essential: Shifting away from headcount-driven metrics, AEC firms should explore skin-in-the-game approaches—like bonuses for outcomes, equity stakes, and public-private partnerships.
  5. Client and Product Manufacturers Must Both Change: Building-product innovators report frustration as AEC professionals and owners often resist new solutions. KP encourages the industry to break the cycle of “the way we’ve always done it” and fully explore prefabrication, modularity, and automation.

All this and more on this episode of the Zweig Letter podcast.

KP Reddy is the founder and CEO of Shadow Ventures and a recognized voice on innovation in architecture, engineering, and construction. With expertise spanning AI, robotics, and automation—plus his role as a lecturer at Georgia Tech—KP brings practical strategies that push the AEC industry forward.

Links referenced in this episode:

  1. Shadow Ventures
  2. KP Reddy on LinkedIn
  3. "Creating the Intangible Enterprise" by KP Reddy
  4. Zweig Group & ElevateAEC Conference
  5. Learn about the Zweig Letter and subscribe: https://thezweigletter.com/
  6. Connect with Randy Wilburn on LinkedIn

Get your FREE Subscription to the Zweig Letter Newsletter.

Stay tuned for more enlightening content from the Zweig Letter podcast, and make sure to subscribe for regular updates!

Other episodes you'll enjoy:

Architecture with Heart - Carley Chastain

From Specs to Stories with Cherise Lakeside

Bridging Design and Construction with Dan Crist

AI Transforming AEC with KP Reddy

Connect with Zweig Group:

Instagram: Zweig Group

Facebook: Zweig Group

Twitter: Zweig Group

LinkedIn: Zweig Group

Website: Zweig Group

Transcripts

Speaker:

Welcome to the Zweig Letter Podcast. I'm Randy Wilburn.

Speaker:

Today's episode comes from a special series we recorded

Speaker:

at Zweig Group's Elevate AEC conference in San

Speaker:

Antonio, Texas. For those who aren't familiar with

Speaker:

Elevate AEC, it's Zweig Group's annual

Speaker:

gathering of AEC industry leaders.

Speaker:

3 days of keynotes, breakout sessions,

Speaker:

awards, and candid conversations about the future

Speaker:

of architecture, engineering, and

Speaker:

construction. The content we captured was exceptional,

Speaker:

and we wanted to share it with a wider audience.

Speaker:

So whether you're listening to a keynote address or a breakout

Speaker:

session, you're getting insights from the people shaping the

Speaker:

industry. And if you want to learn more about attending a

Speaker:

future Elevate AEC conference, visit

Speaker:

zwieggroup.com. Now

Speaker:

let's dive in.

Speaker:

Welcome to the Zweigletter Podcast, putting

Speaker:

architectural engineering planning and environmental

Speaker:

consulting advice and guidance in your ear. Zweig

Speaker:

Group's team of experts have spent more than 3 decades

Speaker:

Elevating the industry by helping AEP and

Speaker:

environmental consulting firms thrive. And these

Speaker:

podcasts deliver invaluable management, industry,

Speaker:

client, marketing, and HR advice

Speaker:

directly to you free of charge.

Speaker:

The Zweig Letter Podcasts, elevating the design

Speaker:

industry one episode at a time.

Speaker:

Okay, welcome to our speed business luncheon

Speaker:

where we are going to focus on innovation, and I mean real

Speaker:

innovation, not the LinkedIn version that so many talking heads

Speaker:

are carrying on about, not the core values version where so

Speaker:

many firms call themselves innovative yet have very little

Speaker:

proof of any real innovative practice. Real innovation

Speaker:

is the creation and implementation of new ideas,

Speaker:

products, services, or processes that deliver

Speaker:

meaningful value by solving problems or meeting needs

Speaker:

in ways that are significantly better than existing

Speaker:

alternatives. It is really the antithesis of "this is

Speaker:

the way we've always done it." Our lunch program is brought to you by

Speaker:

the person and organization I consider the foremost

Speaker:

authority on real innovation, KP Reddy. Now

Speaker:

KP is the founder and CEO of Shadow Ventures, a seed-stage

Speaker:

technology investment firm, and KP Reddy Company, which was formerly

Speaker:

Shadow Partners. He is a globally recognized

Speaker:

authority in AEC environments, AI, robotics, automation,

Speaker:

mobile applications, and cloud computing. KP is a civil

Speaker:

engineer by background, is a frequent lecturer at the Georgia

Speaker:

Institute of Technology, and is a sought-after subject matter

Speaker:

expert. Most recently, KP's experience with AI led him

Speaker:

to publish his latest book, Creating the Intangible

Speaker:

Enterprise. His keynote on the future of innovation

Speaker:

in AEC should help us envision a future where the speed of

Speaker:

business accelerates and AEC becomes an even better

Speaker:

investment. So please help me welcome KP Reddy to the Elevate

Speaker:

stage. You tell Chad you love Metallica and you get to listen to Metallica a

Speaker:

bunch. So

Speaker:

I'm going to tell you a little bit about like what we've been working on,

Speaker:

and I think this this narrative keeps coming up, right? I heard it in the

Speaker:

last panel with my colleague Frank, which is what about our

Speaker:

clients? What about the market? And it's interesting

Speaker:

that because our primary form of communication with our clients is

Speaker:

RFPs and responding to RFPs, we never get to ask

Speaker:

a lot of great questions outside of a procurement

Speaker:

cycle. So what we really looked at

Speaker:

was Let's go spend time with owners. So I

Speaker:

went and spent time with a bunch of your customers, lots of large

Speaker:

owners, municipalities, DOTs, etc., and we spent a bunch of

Speaker:

time talking to them about what's working and what's not

Speaker:

working. The resounding feedback from them

Speaker:

is, we hate working with the AEC industry.

Speaker:

They are a pain in the butt. They tell us what to do.

Speaker:

We are not the customer. We feel— they make us feel like we're the vendor.

Speaker:

They don't add a lot of value. They're a necessary evil. You guys

Speaker:

believe you're a necessary evil? That's what they

Speaker:

think. So we talked to a bunch of owners, and what they said was

Speaker:

all this technology— and I always joke around, I wrote a book on BIM about

Speaker:

13 years ago. Nobody should be buying it, yet I still have

Speaker:

sales because BIM is dead. No one cares about BIM.

Speaker:

That's an old construct. But what they said resoundingly is

Speaker:

nothing has changed. All this technology, all this

Speaker:

innovation, at the end of the day, our projects are not on

Speaker:

time, they are not on budget, and the quality actually is a question

Speaker:

mark. So why are we doing all this? And one

Speaker:

of my colleagues did a deep study on the design and building of the

Speaker:

Empire State Building, and his conclusion was

Speaker:

in today's age we could not meet the metrics, we could not meet

Speaker:

cost, we could not meet delivery schedules, that they did

Speaker:

it better back in the '50s than we're actually doing it now.

Speaker:

So that's the premise. So we did about 101

Speaker:

interviews. We created a thing called the Integrated Owners Forum. We got all the top

Speaker:

owners in the world to start showing up and just ask very

Speaker:

free-form questions, everything about How do you feel about hiring

Speaker:

architects? How does that process work for you?

Speaker:

How do you hire contractors? What do you think about bid? What do you think

Speaker:

about CM at risk and design-build? Asked them all the questions

Speaker:

and said, you are the customer. We need to

Speaker:

serve you the way you want to be served. And that was like, it was

Speaker:

weird because they said, no one's ever asked us that.

Speaker:

No one's ever asked us what we want. We've been told

Speaker:

draft an RFP, put a bid document together, do your plans

Speaker:

and specs. And so I just kind of said, let's get back

Speaker:

to first principles. How do we want to do things different? Let's

Speaker:

stop iterating on an old construct and let's just start

Speaker:

over. So we started to really explore their thinking, and my team

Speaker:

spent a lot of time on this stuff. I wish they

Speaker:

would have used AI, but they'd spent way too much time.

Speaker:

One of the categories in here that we spent the most time with was actually

Speaker:

some of the building product manufacturers that said there's a massive

Speaker:

disconnect between how they think about innovating the next HVAC system,

Speaker:

the next electrical distribution system, and their connectivity to the

Speaker:

people that design and specify with their products, that there's

Speaker:

a massive disconnect. In fact, they said that it's hard for them to

Speaker:

innovate because they have to convince you to specify it. They have

Speaker:

to convince you to design with it, and it's not worth it. Like, why

Speaker:

innovate something new if the industry's not gonna adopt something

Speaker:

new? And the owners don't know better, right? Some owners do. When I'm dealing

Speaker:

with Intel or Genentech or Meta and all that, they're a little bit more

Speaker:

savvy, but the local K-12 system doesn't

Speaker:

understand what the latest system is from Schneider Electric.

Speaker:

So this was kind of our nutshell, and anyone that wants this report,

Speaker:

you just email me, kp@kpready.co, I'll have someone send it to you.

Speaker:

But this is what we found out. Turns out owners admitted

Speaker:

to us, they don't know what they're doing. They're like, we, just

Speaker:

like you guys can't hire the right people, can't find the right people, guess what?

Speaker:

Neither can we. We go build a, a

Speaker:

manufacturing facility for 3 years, we burn out our people,

Speaker:

then they quit. We rarely have people on our side stay for

Speaker:

one more— more than one capital program because it's

Speaker:

brutal. It's a hard business. And so they end up quitting and they

Speaker:

go somewhere else. So they admitted we don't have the

Speaker:

capability. They're like, you know, budgets and schedules are

Speaker:

just kind of novelty. We don't really

Speaker:

understand what's going on. The other thing is they said, look, as much as we

Speaker:

spend time on design and thinking, we have this problem where

Speaker:

there's two silos. There's the CapEx buyer and there's the OpEx buyer,

Speaker:

that the industry designs and builds buildings and they

Speaker:

can't tell us what the operating costs will be. That's

Speaker:

like going out and buying a car, not knowing what the maintenance costs

Speaker:

will be, not knowing what your insurance costs will be, not knowing what your miles

Speaker:

per gallon are. And they're like, we do this with billion-dollar purchases,

Speaker:

we are going in blind. Practically criminal if you ask

Speaker:

me. So the data aspect, they're like, all this

Speaker:

data is in silos, nobody seems to care our

Speaker:

opinion on the data, it's our data. By the way, every one of

Speaker:

our owners believes what you guys think is your data, they believe

Speaker:

is their data. I'm not gonna argue, that's up to the lawyers.

Speaker:

But that is their belief system. So you can either say, hey, we're not gonna

Speaker:

share data and argue about it, or you

Speaker:

can say in this transparent world, you say, actually it is your data.

Speaker:

In fact, we're happy to help you monetize that data.

Speaker:

We'll help you use it better. We'll make it usable for you. Tell us what

Speaker:

you wanna do with the data. And I thought that was pretty fascinating.

Speaker:

This whole idea of BIM, all we're doing is rehashing 2D into 3D.

Speaker:

There is no data and nobody cares. There's so much

Speaker:

contractual misalignment. In other words,

Speaker:

we get paid a percentage of the project. If the project

Speaker:

gets increases in scope and price, we actually get paid

Speaker:

more. So there is no benefit to delivering cheaper, better,

Speaker:

faster to the AC industry. There's no bonus involved.

Speaker:

And so we saw like massive, massive misalignment and there's

Speaker:

misalignment and there's actually, um, really

Speaker:

opposite alignment, right? In other words, the benefit of

Speaker:

a bad project actually might benefit the AEC industry

Speaker:

more, uh, if it goes poorly financially.

Speaker:

Once again, this is what they said. We asked for specific

Speaker:

instances, they shared them with us. So I'm not saying it's 100%,

Speaker:

but we have to remember there's facts and there's feelings. These are

Speaker:

feelings that are rooted in facts, right? They've had these

Speaker:

experiences. The lack of information,

Speaker:

can't get as-builts, nobody seems to be working for us.

Speaker:

One of them said, I love working with architects that talk about the building being

Speaker:

their building. There's like, it's not our building. You didn't write a billion-dollar

Speaker:

check. It's not your building, sir. I paid. And then

Speaker:

how to communicate. Which is even becoming harder. Oh, let's set up

Speaker:

Procore, let's do this, let's do email. You know, it's gotten

Speaker:

worse and worse. In fact, all this technology that we have—

Speaker:

how many of you feel like we're communicating better

Speaker:

right now? We have all these Zoom notes flooding our

Speaker:

inbox that we don't have time to read, and

Speaker:

everybody's dropping it into AI to hopefully make some sense of it. So

Speaker:

this is kind of what they came up with which we found extremely

Speaker:

fascinating. So, this was kind of

Speaker:

the conclusion, right? So, we said, okay, great, let's put our analytical

Speaker:

hat on and said, we know it's broken, where are some of the choke

Speaker:

points? And a lot of it came down to

Speaker:

scope. It turns out your customer doesn't know what they

Speaker:

want. And it turns out we tend to ask

Speaker:

them a series of questions and iterations along the way, to help them define what

Speaker:

they want. And it was funny, as they admitted it, they're like, yeah, we're, we're

Speaker:

a big part of the problem. We think we know what we want, but we

Speaker:

don't actually. And so that iteration that we go through with the

Speaker:

AE firms costs us time and money. And, you know,

Speaker:

when our boss comes and yells at us about why things are taking too long,

Speaker:

well, we blame the AE firm. It's easy enough, right? But they did

Speaker:

admit that they don't actually know what they want. So the idea of

Speaker:

developing better requirements, deeper requirements,

Speaker:

in a detailed way and understanding scheduling. And, you know, we

Speaker:

used to joke around. I used to do

Speaker:

manufacturing facilities, so we'd have a job trailer full of

Speaker:

P6 printouts of the schedule, right? And we always used to

Speaker:

joke around, the minute we print it, it's wrong, right? And

Speaker:

so a lot of what they're asking for on the scheduling is not so much,

Speaker:

you know, what's the schedule, but what are the variables? What's going to hurt us?

Speaker:

On the schedule. And guess what? There's 3 different AI

Speaker:

weather sources now. We have so much data that we can work with

Speaker:

to actually start extrapolating risk and schedule, risk and

Speaker:

budgets, and things like that. The firm

Speaker:

selection— how many of you, how many of you like responding to RFPs,

Speaker:

right? It's a lot of fun. Now we have AI doing it for us. That's

Speaker:

going to make it better. Um, how many of us actually believe

Speaker:

that They're hiring and selecting based on QBS,

Speaker:

on qualifications, right? So the clients, the owners started

Speaker:

to understand like, hey, we're actually part of the problem, which I thought was great.

Speaker:

It was very big of them to admit that they were part of the problem.

Speaker:

So that's the thing that we think is an opportunity with AI. And

Speaker:

I'll get into like why AI even matters with any of this

Speaker:

stuff.

Speaker:

A lot of the overall iteration of design and them not

Speaker:

understanding what decisions are being made. They don't know what they're

Speaker:

signing up for. We say, hey, we think we should go this direction. They say,

Speaker:

okay, well if that's what you think, you're the expert. What they didn't tell you

Speaker:

is like, oh, that's probably gonna cost us more in energy costs, that's gonna cost

Speaker:

us more in maintenance, right? They don't necessarily understand the

Speaker:

downstream implications of any decision they make. Then

Speaker:

of course, everybody loves to blame the contractor. Some of my best

Speaker:

friends are contractors. Once again, it's back to

Speaker:

lack of information, lack of information.

Speaker:

Everybody know this thing? Why does this

Speaker:

exist? Why does the iteration of design

Speaker:

exist? Why don't we design on day 1 to

Speaker:

shop drawing level? Right? We do this

Speaker:

because we started off sketching on vellum and

Speaker:

drawing, and the cost of change to make any change

Speaker:

was so expensive that we decided like we need to

Speaker:

iterate, which made a lot of sense, right? Let's iteratively go through the process with

Speaker:

the customer, make iterative changes, make iterative

Speaker:

decisions, avoid a lot of rework. I was working on a big project

Speaker:

and my client was a big structural engineering firm. And I went to a meeting

Speaker:

with them, and the architect said, so where are we

Speaker:

on the design, on the structural design? Structural engineer said, yeah, we're on track. Didn't

Speaker:

show him anything. And I said, like, do you have anything? Like, oh, we

Speaker:

haven't even started. And I'm like, why haven't you started?

Speaker:

If we made design changes every time the architect and owner

Speaker:

changed their mind, we would be out of business. So we just kind

Speaker:

of sit in meetings and wait till they iterate whatever it is they want to

Speaker:

iterate. And we just wait till they get done with their noise, and

Speaker:

then we start designing. Because if we design on the first meeting, man,

Speaker:

we'll lose all our money on these projects. And I found that pretty

Speaker:

fascinating, right? So if you think about it, we've gone

Speaker:

through this cycle because the rate of change,

Speaker:

because we lacked a lot of technology, was so slow.

Speaker:

Making a change meant let's print out another set of drawings.

Speaker:

Go, uh, you know, sketch out and

Speaker:

erase on paper, on vellum. Then we went to CAD, and of course that's— for

Speaker:

some reason we thought the rate of change was also very difficult. But a lot

Speaker:

of this is we don't have enough information,

Speaker:

right? We're also feeding real-time information from the customer, from

Speaker:

codes and everything else. And as we get further along, we get

Speaker:

more and more information. And so you think about it, we, we draw a

Speaker:

box and then we start coloring in little pieces of it and drawing in the

Speaker:

pieces of it, and that takes time and energy. So

Speaker:

this is kind of where we've been, right?

Speaker:

So I did a small stint, 18 months, with Frank Gehry to run his

Speaker:

technology group. Frank had a belief system that you did

Speaker:

drawings to a shop-level drawing,

Speaker:

and his point was the dumb contractor won't know what to

Speaker:

do with my designs. So if we don't tell them exactly what to do,

Speaker:

the project will go over budget. And so in his

Speaker:

team, we had material scientists, we had structural engineers, we had lots of people, but

Speaker:

we delivered a full set of shop drawing level

Speaker:

drawings to the contractor. A lot of people told him like, hey, that's a

Speaker:

lot of liability, which we can talk about a little bit. There is

Speaker:

definitely a narrative whenever someone a customer asks you for something,

Speaker:

if your first response is, "Nope, can't do it, it's

Speaker:

too much liability," versus the answer being, "Yes,

Speaker:

but let me take a look at it and let's see how we can capture

Speaker:

that liability. Maybe we can do a joint capture on the liability,

Speaker:

but let's understand what the implications of it is." That was one of the biggest

Speaker:

feedback that the owners gave us, was that our industry says

Speaker:

no a lot. They tell us they can't do stuff, and it's not that they

Speaker:

can't do stuff, it's that they don't want to do stuff. We also asked them

Speaker:

questions like, would you be willing to pay more? They said absolutely. If they can

Speaker:

start demonstrating their methodology, how they create value,

Speaker:

how it'll impact the project, happy to talk to them about a

Speaker:

different fee structure. The funny thing is they're like, you know, we have to— we're

Speaker:

the customer and we have to go buy AIA contract

Speaker:

docs. It's like, isn't that fascinating? We sign

Speaker:

our vendor's contract. We're the

Speaker:

customer. They should be signing our contract. They should work the way

Speaker:

we wanna work. And you're starting to see that with some of the superscalers

Speaker:

that have realized that I don't wanna operate under the

Speaker:

existing construct. I'm gonna develop my own way, right? When Elon Musk, whatever you

Speaker:

think about him, decides to build a data center in Memphis and just

Speaker:

breaks all the rules and does whatever he wants and delivers a data center in

Speaker:

9 weeks, One part of the narrative is like, wow, he got

Speaker:

something done in 9 weeks that the industry would've never done in 9 weeks.

Speaker:

Other people say, well, he broke all the rules. We have to live by the

Speaker:

rules. He got to break all the rules and he's paying a million dollar a

Speaker:

day fine for breaking all the rules. Funny thing is

Speaker:

he doesn't care about the million dollar a day fine. It's built into the business

Speaker:

model. So where are we going from this and how do we start to think

Speaker:

about it? So Frank asked a question

Speaker:

about agentic AI. I'm not gonna get too deep into agentic AI.

Speaker:

I'm, I'm around all day today. If anyone wants to talk about agentic

Speaker:

AI, we can talk about it. But this is how we see the

Speaker:

future of what we're doing, that everything is kind of tech-enabled,

Speaker:

a lot more integration. Um, I talked yesterday about some of the

Speaker:

things that I've built. One of the things that I wrote up was kind of

Speaker:

a 10-page academic paper that I gave to all these

Speaker:

VCs. Which was the death and

Speaker:

reemergence of the master builder. And like, why did the master builder go

Speaker:

away? What was the construct of that? And how do we bring

Speaker:

it back? Because the net result of that

Speaker:

was positive. So as we think about how do we

Speaker:

reemerge as the new master builder as an industry and get

Speaker:

rid of the silos, these are some of the key points that we

Speaker:

really believe. I you say, know, it's funny Chad's talking about LinkedIn.

Speaker:

I'm very active on LinkedIn.

Speaker:

Um, on my profile it says AI design buildings,

Speaker:

robots build buildings. My timeline is

Speaker:

5 years. My timeline is 5 years.

Speaker:

I might be wrong on my timeline. I will not be wrong on the end

Speaker:

state. The end state is the right state. And a lot of it's because of

Speaker:

what we're talking talking to our owners about. We have owners, we have

Speaker:

some companies that do fully automated drywall and painting and

Speaker:

finishing. When I tell the owners about that, they're like, nobody told us that,

Speaker:

nobody told us that. So as

Speaker:

owners continue to get educated and exposed to these technologies,

Speaker:

they're going to start asking a lot of questions.

Speaker:

So how do we start looking at that? The entire model, the software

Speaker:

model— I know there's a lot of software vendors out there. We kind

Speaker:

of believe software is kind of dead as a way to go

Speaker:

buy software. It's really more software is

Speaker:

just embedded in how we do things. We think that a

Speaker:

lot of this is shifting just how people think about executing work.

Speaker:

I work with some very large subcontractors that invest heavily in R&D

Speaker:

to deliver a different way. And so we think this is how

Speaker:

The shift away from headcount as a growth metric to more like, how do we

Speaker:

think about project delivery? How do we think about

Speaker:

project profitability? I'm working with a large NFL team. I mean, they're

Speaker:

all large, but on a new stadium. And I said like,

Speaker:

hey, as a consultant, can we help you do better? Can we guarantee the outcomes,

Speaker:

the budget, the schedule? And if we guarantee it, I want a cut.

Speaker:

I'm like, oh, you want a bonus? I'm like, no, I want to own part

Speaker:

of the team, man. Like, that's what I want. He's like, well, I can't let

Speaker:

you in part of the team. However, I could give you a bonus if you

Speaker:

can help us do it faster. So the thinking around even the financial

Speaker:

models and the business models is even starting to change

Speaker:

to more of a developer mindset, to more of a skin-in-the-game

Speaker:

mindset. I think, you know, someone's asked me a couple of questions about

Speaker:

DOTs and how do I see that going. There's a reason why

Speaker:

we're seeing a lot of public-private partnerships. There's a reason

Speaker:

for that, and those are massive, massive economic opportunities for the industry.

Speaker:

Unfortunately, the asset allocator is probably getting

Speaker:

paid on equity and upside, and they're going to pay you hourly

Speaker:

rate, and they're going to try to negotiate your rate down and

Speaker:

not give you a piece of the action. Because I'll tell you what, long term,

Speaker:

in terms of how you think about growth and economic growth, owning

Speaker:

part of the asset is probably a better move, and these

Speaker:

developers are not going to let you in on the action.

Speaker:

So how do we think about this? If you imagine, which

Speaker:

you don't have to imagine, we're actually building a lot of these. How do you

Speaker:

think about getting information out of the owner, out of the

Speaker:

customer, right? We ask a series of questions, we maybe

Speaker:

have some process, they give us some feedback, we go away, We

Speaker:

analyze that, then we have more questions and we go back to them.

Speaker:

What if instead the requirements process was much

Speaker:

more conversational with an AI bot based on all your information?

Speaker:

So you could have a 1-day meeting with a customer and

Speaker:

actually understand exactly what they need.

Speaker:

Customer says, hey, I wanna build a 350-bed hospital.

Speaker:

Fantastic. What are you thinking about surgical centers? Oh, I hadn't thought

Speaker:

about that. Well, I think we need 5. Okay, great. What's your typical

Speaker:

footprint for a surgical center? What type of equipment? Oh, we're

Speaker:

going to put an X-ray machine in there. We're going to put a robotic surgery

Speaker:

thing. Great. Which model? Right? Because the AI has access to all

Speaker:

this information on the front end and can continue to prompt

Speaker:

the customer real time with— to answer all the

Speaker:

questions. At the end of this, now you actually know the requirements.

Speaker:

You've actually learned a lot about what the customer needs, which

Speaker:

means you can push all of those requirement gathering instead of being iterative,

Speaker:

iterative, and along this entire long process, we're

Speaker:

getting as much information upfront as possible. This is something we've already

Speaker:

done, right? So we've already looked at this.

Speaker:

How do these requirements fit into design, right?

Speaker:

Design is about taking a lot of set of variables from the customer

Speaker:

And the business case, this is another interesting thing. We ran a model where

Speaker:

we looked at a new hospital and we looked at staffing and the revenue model

Speaker:

and the big business plan for the new hospital, and we fed it into

Speaker:

our system to understand how that would inform design.

Speaker:

What would be the ROI of that new surgical center? Do you actually need more

Speaker:

surgical centers based on the demographic data you're giving us, based on the

Speaker:

certificate of need data? So we're pulling all this data and the owner's

Speaker:

saying, Wow, it turns out you're right, we do need an extra surgical

Speaker:

center. Our fancy consultants at McKinsey missed that,

Speaker:

right? So this is the type of iteration you're having up front. So if you

Speaker:

think about, if you have all the requirements up front,

Speaker:

you now all have all the data to push into the

Speaker:

agentic design. So now you're designing very quickly on

Speaker:

the fly. And guess what? If the customer changes their mind, That's

Speaker:

fine, it's quick, but we're also designing the shop drawing

Speaker:

level. We're also designing with an idea that is the product available?

Speaker:

We typically specify this HVAC system, we're pinging

Speaker:

the manufacturer to say, does that exist still? Are the performance

Speaker:

specs still the same? Is it gonna be available based on the

Speaker:

predicted schedule of my project? Will it be available? 'Cause if it's

Speaker:

not available, which if you're dealing with the superscalers, that's one of the biggest

Speaker:

problems is Product availability, right? Systems availability.

Speaker:

We're actually designing and getting real-time data to inform the design and

Speaker:

make those changes. Why do we have RFIs,

Speaker:

requests for information? Because we didn't provide enough information.

Speaker:

We didn't give enough prescriptive information. So that's what we're

Speaker:

starting to see there. Tech-enabled construction services. Is it gonna be

Speaker:

robots on the job site? Not one of them. But

Speaker:

we're already thinking about how do we do orchestration? What does the superintendent of the

Speaker:

future look like? And is it a set of cameras that are

Speaker:

actually sending data to all the different robots on our job site? I

Speaker:

think I talked about it yesterday. We have an automated grading system,

Speaker:

right? Grading equipment. We now have automated painters and drywallers.

Speaker:

We also have prefab that's coming. So not everything's gonna be on the

Speaker:

job site. And you have a lot of other interesting things that are being built

Speaker:

every day. The cost of robotics are getting cut like

Speaker:

5% a month, and it's just getting cheaper and

Speaker:

cheaper. So we're going to have access to the type of robotics that only

Speaker:

very large industries have had in the past, like automotive.

Speaker:

Autonomous construction, big part of that, whether

Speaker:

it's prefab offsite or whether it's— the offsite

Speaker:

stuff was super interesting. We talked to a large healthcare provider, and they said, we

Speaker:

tried, we found a company that does prefab patient rooms,

Speaker:

a company called Bloxx, I think, out of Birmingham. But the

Speaker:

problem is we didn't talk to them until after

Speaker:

we realized we may not meet the budget and the schedule.

Speaker:

And they said, we still need lead times. Why didn't your architect design

Speaker:

with us in mind? Well, because the

Speaker:

architect wanted to design the rooms they wanted to design the rooms instead of thinking

Speaker:

kit of parts. So now that owner now realizes like, why am I—

Speaker:

when I start a project day one, I'm going to look at who builds prefab

Speaker:

patient rooms and I'm going to force my architect to design with

Speaker:

modularity in mind. And you're seeing this across the board. It's not just

Speaker:

healthcare, it's a lot of different markets. Actually just yesterday, Schneider Electric

Speaker:

and Compass Data Centers announced their

Speaker:

pre-manufactured data center in a box. So between Schneider and

Speaker:

Compass, they're just gonna ship you a data center, right?

Speaker:

You're also seeing it with other, in other markets where they're realizing that prefab

Speaker:

and offsite is really important. The last piece that I think

Speaker:

is really important to customers is facilities management.

Speaker:

Great, you guys built the house, you designed it, you built it, you left. Now

Speaker:

I have to live with it for the next 20 to 50 years.

Speaker:

We're working with a large hospital project They called us in and they

Speaker:

said, we met— we hired one of the top architecture firms. They

Speaker:

seem to fail to realize that within 5 years we'll probably have robots in

Speaker:

our hospital. And they didn't design with the idea of

Speaker:

coexistence of robots, humanoids or otherwise,

Speaker:

in the space living with humans.

Speaker:

Seems like a big miss since we probably won't tear down this hospital for 50

Speaker:

years. It's not going anywhere. Can we all agree that within 50 years we'll have

Speaker:

robots working with us? So you have to design a building that has robots

Speaker:

working with us, that has that optionality.

Speaker:

So why, why build these types of service platforms? So as you think about

Speaker:

AI, once again, not a feature, not a tool

Speaker:

necessarily, part of an equation. How do you think about

Speaker:

building your AI service platform and rethink your

Speaker:

business model? Because guess what? If you have the right conversations with your

Speaker:

your customers, they want a different business model.

Speaker:

I'm gonna skip through on time. So a lot of what we're looking

Speaker:

at is we've built decision to support systems, we're

Speaker:

building requirements gathering systems, prefab,

Speaker:

agentic design and systems to do the entire lifecycle.

Speaker:

It's early innings, but we're still, we're on

Speaker:

path. So, I think I'm out of time.

Speaker:

One of the things we do for the industry

Speaker:

and one of the things we really start to think about is

Speaker:

how do you assess where you are and where you're going? I posted something a

Speaker:

couple weeks ago on change management. The change management is made up,

Speaker:

it doesn't apply. Change management implies that you know what the end state is.

Speaker:

We do not know what the end state is.. And if you

Speaker:

constantly focus on managing to an end state that is probably fictional,

Speaker:

you will probably lose. So a lot of like even what Frank was talking about

Speaker:

up here earlier is really about how you think about

Speaker:

assessing where you are and where you're going. Thank you very much. Hopefully

Speaker:

I didn't ruin your lunch. Apparently I did that for a

Speaker:

couple people yesterday. But with that, we're going to move

Speaker:

on to our AAC Innovator Awards.

Speaker:

I think they're going to cue something up. Thank you, everybody.

Speaker:

Thanks for tuning in to the Zweig Letter podcast.

Speaker:

We hope that you can be part of elevating the industry and that

Speaker:

you can apply our advice and information to your

Speaker:

daily professional life. For a free digital

Speaker:

subscription to the Zweig Letter, Please visit

Speaker:

theZweigLetter.com/subscribe to

Speaker:

gain more wisdom and inspiration in addition to information about

Speaker:

leadership, finance, HR, and marketing your

Speaker:

firm. Subscribe today.

Links

Chapters

Video

More from YouTube