December 2, 2025 • 55 mins
Mike Sewell joins the episode to discuss his data-driven approach to environmental design and the role of technology in the AEC sector. He shares insights into his career and motivation, emphasizing the importance of improving communities through design. The conversation highlights the Empath platform, its technical origins, applications, and pilot projects, demonstrating its business impact and cost savings. Mike addresses privacy and compliance in design, stress factors in transportation, and potential applications in healthcare and construction. The discussion also covers automation's impact on AEC, showcasing innovations to clients, and technology's role in hiring and employee retention. Mike concludes with advice for the younger generation.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
I've got a really nice route.
(00:01):
I looked compared started comparing ridesbetween The Netherlands and The United States,
and I had way more spikiness in my heart rate.
And there was elevated levels of heart ratewhere it shouldn't be.
Like, it was a very comfortable ride.
And I was like, there's something behind thisdata.
And that that was what sparked, hey.
I could probably use this to do a calculationand then have people share their data and and
(00:23):
target places that need to be improved.
So that that's the origin story, and I I lovethat.
Have you ever wondered how successfularchitecture, engineering, and construction
companies scale their business?
Or have you ever wanted guidance on how to getmore growth, wealth, and freedom from your AEC
company?
Well, then you're in luck.
(00:45):
Hi.
I'm Will Forrette.
And I'm Justin Nagel, and we're your podcasthosts.
We interview successful AEC business leaders tolearn how they use people, process, and
technology to scale their businesses.
So sit back and get ready to learn from theindustry's best.
This is
building scale.
(01:07):
Hey, listeners.
It's Will here.
Our mission is to help the AEC industry protectitself by making technology easy.
If you've ever listened to our show, then youknow that the three pillars of scaling a
business are people, process, and technology.
So if you suspect technology is your weak link,then book a call with us to see where we can
help maximize your company's IT cybersecuritystrategy.
(01:30):
Just go to buildingscale.net/health.
Today's guest is Mike Sewell.
Mike is the director of innovation and a boardmember at Gresham Smith, one of the nation's
leading architecture, engineering, andconsultancy firms.
With over two decades of experience, Mike isknown for pioneering safety focused design and
leveraging emerging technologies to improveinfrastructure for vulnerable road users.
(01:55):
He's the creator of the patented empathplatform, the first of its kind to quantify
emotional response in different environments.
Empath earned honors, major honors, includingFast Company's world changing ideas award and
architect magazine's r and d award.
Mike also serves on the board of the league ofAmerican bicyclists and has testified before
(02:15):
congress on multi module safety andinfrastructure funding.
His work at Gresham Smith continues to blendtraditional engineering with cutting edge
innovation, all aimed at creating healthier,safer, and more connected communities.
And with all that said, Mike, welcome to theshow.
Will and Justin, I really appreciate you,affording me the opportunity to come and speak
(02:37):
about something I love, which is this weirdkind of mashup of design and technology.
So I'm I'm really thrilled to be here.
No.
It's super fun.
I I get giddy, thinking about what we're gonnatalk about because the empath thing is is wild,
and what you're doing as an innovation leaderis just fun.
It's just totally down our our pipe.
(02:58):
And it totally tickles the funny bone when youthink about, like, you know, there's a lot of
talk about AI, but this this is this isinsanity.
This is a place that I did not fathom that itwould go, and yet here we are.
So don't wanna keep you waiting.
So Justin, why don't you why don't you get usstarted?
(03:18):
Yeah.
So, Mike, I said all these wonderful nicethings, but tell us your origin story.
How did you get in engineering?
How did you eventually get into the innovationpiece of that?
And, tell us a little bit more about GreshamSmith.
Yeah.
Sure.
Well, in the intro, I might need to prop up myhead for a second.
That was very kind of you.
Thank you for going through all the accolades.
But, honestly, it's it's been a labor of love.
(03:38):
Started with with a background in engineering.
I real I chose engineering because, you know,when I was growing up, we were very reliant on
a lot of public public outreach, public thingsthat were not in place that really made our
jobs a lot a lot more difficult than I thought.
If I chose engineering, I could I couldpotentially help the communities that I love,
and and prop up other folks.
(03:58):
So that's the reason I gravitated towardsengineering, which has been a really fulfilling
career in and of itself.
But then the tech side, I'm a dork.
You'll figure that out over the course of thispodcast.
We're pretty pretty you know, you'll figurethat out pretty quick.
But, I've always been just enthralled withtechnology.
So going back as far as, the mid eighties whenI got my first home computer that was, you
(04:21):
know, gifted to the family, I I taught myselfeverything, autodidact all the way.
Like, never have taken any, computer courses,just have poured myself into finding out and
staying curious about it.
And so I got a lot of background with that too,but then I had the opportunity, at at Gresham
Smith to to lead the innovation program, whichis kinda wrapping that technology side and that
(04:44):
love for technology around the the genuine carethat I want, as an engineer and and now
architects as well.
So it's just it's that's a little bit of backabout the background.
I'm happy to go into more detail, but that'sthat's the origin story.
Alright.
Well, let's hit on that core purpose of, youknow, you essentially wanting to make your
community better or your communities better.
(05:04):
Right.
What oh, let's let's dig into that a littledeeper.
What what does that truly mean to you, and andwhy does that drive you?
Yeah.
Well, it's it's pretty simple.
You know, while some of us can choose forms oftransportation, some of us choose, how we
interact with an environment around us, a lotof people don't.
I didn't.
(05:25):
When I you know, when we were growing up, we wehad it rough.
And, you know, now, you know, you you talked alittle bit about my, background with and love
for multimodal kind of engineering andtransportation.
A lot of people, that's not a choice.
Like, that's that's the only viable solutionthat they can afford to get around.
And I think that really solidified.
(05:45):
Those folks don't have a voice, those folks arevery dependent on places that I influence
potentially as an engineer or designer ofplaces, so why not really invest myself in
doing what's right by by the communities thatdepend on better and and really should be
demanding better places and better connectionsto those places.
So it it's as simple as that.
(06:06):
Just looking out for people that depend onplaces and infrastructure.
Alright.
I'm not gonna try to hide the headline here orbury the headline as they say.
Let's dive into the empath stuff.
So what Oh, yeah.
First, what is it?
And second, where did it start?
Where did this idea come from?
Sure.
Yeah.
So in innovation, this kinda kicked off empathat at Gresham Smith.
(06:30):
This was the impetus behind my current role,which was I started really challenging,
engineers like data.
Right?
You you we like a lot of numbers.
We like a lot of supporting information, andthere was a gap.
I think in general terms, I'll pick onengineering, but but one thing in particular
always bug me about engineering is we wait alot of times to challenge the status quo until
(06:51):
something catastrophic happens.
We we wait until someone loses their life.
We wait until someone has a serious injury, andthen we're like, oh, we should probably take a
look at this roadway or we should take a lookat this space that's caused, this problem.
The problem is, like, that's real life.
That's that's a that's somebody that that hadto be on a piece of infrastructure or in a
(07:11):
place and something, unforeseen happened tothem.
That's a terrible data point to wait for.
And so that was the spark, behind Empath.
What it is is it looks for precursorinformation.
It's I will dork out for a second, but it'sit's using low polling heart rate data on any
modern wearable or Bluetooth device.
(07:32):
It is looking for nonlinear transforms of heartrate.
If you look at the two calculations primarilythat calculate stress, like if you go into a
stress test, it's it's using EKG level data,which drains batteries.
Mine uses low frequency heart rate data, so youcan record all day long everywhere you go using
using not a lot of battery, and it marries thatdata up with the location information.
(07:58):
K.
So what that in of itself, it's looking forstress.
It's it's calculating fight or flightmechanisms.
So think of it if you walked out and got buzzedby a bus on on the side of the on the side of a
road.
You're not losing your life.
It makes you very uncomfortable.
That shows up in data now.
And so comparing that to the original EKG, thetwo primary methodologies for calculating
stress, they're just they're just slopeanalysis.
(08:20):
And so I dumbed it down and created a novelcalculation for stress.
I had it vetted with Vanderbilt.
And then that that part is fun, in and ofitself, but the real power is aggregation.
So when you've got a lot of people experiencinga similar type of response, well, it's no
longer an individual user.
It's the environment causing that stress.
(08:41):
And so you start to see it's painting a picturenow where if we can go upstream from somebody
having to have a serious injury or or losetheir life and we start looking at these
leading indicators of people not feelingcomfortable in space.
Well, let's go fix that space before, you know,somebody does have a problem.
That's the background.
That's what it is, but the the origin story iseven cooler, I think.
(09:04):
Again, I'm a bike guy.
I I was invited as part of my infrastructurerole to audit, a group of people biking across
The Netherlands.
So I was in The Netherlands, biking across thecountry.
We we biked in miles.
It's about 220 miles across the com thecountry, and I was just, like, hit.
I was outside of Leiden.
(09:25):
I'll never forget it, and we were in this biglong group of people biking, and it was just
this wave of appreciation for how they puttheir infrastructure together hit me.
And I was it it was just magic.
It was beautiful.
It was intuitive.
It was comfortable.
There was plenty of accommodations.
You know, I knew exactly where I was.
Everything just lined up and checked all theboxes for places and for infrastructure.
(09:49):
And I remember I pulled over.
I took a picture of that spot, and I was like,I wanna calculate this.
And so then I, you know, I've mentioned I'm adork.
I I record everything.
I've got I've got sensors on me at all timeswhen I'm on my bike rides.
It know it knows where I'm at, and I startedlooking at the data.
When I got back to the stateside, I startedreally digging into a normal commute.
I used to bike commute every single day.
(10:10):
Just love love that mode of transportation.
And just even when you you feel like, hey.
I've got a really nice route.
I look compared started comparing rides betweenThe Netherlands and The United States, and I
had way more spikiness in my heart rate, andthere was elevated levels of heart rate where
it shouldn't be.
Like, it was a very comfortable ride.
And I was like, there's something behind thisdata, and that that was what sparked, hey.
(10:34):
I could probably use this to do a calculationand then have people share their data and and
target places that need to be improved.
So that that's the origin story, and I I lovethat.
It's just that that hit me there in such apowerful way that I was like, I I know that I
can figure this out, and I'm not, you know, I'mnot a medical.
Medical is not my background.
(10:54):
But, again, engineering is all about data.
It's all about dissecting a problem and tryingto find a new solution.
I know we're gonna talk about digitaltransformation, things like that, but that was,
like, weird, confluence of a couple differentpassions just all coming together at the right
time and in the right place.
So what, what what has come from this?
Right?
So this sounds cool.
(11:15):
Right?
Like, as a listener, it'd be like, okay.
So you understand that certain places are morestressful?
Totally down with that idea.
What has now come out of this?
Yeah.
So a a few things.
For one thought leadership.
I think the demonstration that designers atGresham Smith and I I wanna I wanna give kudos
to the company.
They invested in in me exploring this.
(11:38):
Thought leadership for number one.
We care about the places we design, so whydon't we use intuitive new ways of figuring out
how to do that even better?
That's number one.
We deployed it in places like Denver.
They have a massive population that that lovesmultimodal travel, and and they were they were
really, an early adopter.
(11:59):
So they allowed us to do a pilot test with withDottie out there.
We deployed it.
We had we collected about three and a halfmillion data points across the city.
I I can share graphics of it later, but itreally highlighted areas that they wanted to
target for improvement that there were no crashdata.
So it highlighted again, there's discrepanciesbetween the historic way that we've approached
(12:22):
projects and what the stress data can tell us,an aggregated stress model can tell us that,
hey.
There's a problem here.
Let's let's aggregate all these things.
Let's, prioritize different projects that wethink will have an impact, and we can do quick
fixes to see if that's addressing the problem.
And if it is, then we can invest more money.
So that kind of informed how council wasstarting to allocate dollars.
(12:44):
Lots of other cool things came of that project.
Since then, though, we've also started tryingto take it indoors.
Think of other high stress locations likeairports.
You know, no one thoroughly enjoys the day thatthey have to go through and work out through
security, but the next step for us would belet's use it for a reactive environment.
So let's say you are arriving at an airport andyou say, yeah.
(13:06):
I'm gonna share with a Delta via AggressionSmith my experience journey.
We can map that now.
We can we can highlight specific areas wherethere might be confusion that shows up in
stress data and target those for interventionslike, simple way finding and signage, things
like that.
We've also deployed it I know I'm go I'mrambling now, but it's so there's just so many
different directions.
(13:28):
Think of high stress environments like thehealth care industry, which we design a lot of
of hospitals.
And we're constantly changing, you know, thetypes of environments we create to influence
either, a customer experience or to supportprofessional staff that are in those
environments.
Those people are in really high stress jobs allday long.
So recently, you saw this come to a head duringCOVID.
(13:50):
We had nurses burning out.
Stress.
It's stress.
Mhmm.
So can we start to detect when elevated levelsof stress start to, you know, inject them from
their sweet spot, the the the flow state, youknow, as as people call it.
Let's let's do a mandatory respite in an anAggression Smith designed respite room and get
(14:11):
them back to an optimal level to carry out ajob.
So it started to open our eyes.
There's a lot of other ways that we can applythat kind of approach to positively influence
the design of places.
So I hear I hear the positives.
In terms of, I'm gonna put my business hat onfor a second.
(14:35):
Like a lot of dollars are invested and soundslike, you know, for someone that, you know,
when we talk about the business hat, how doesit impact either I understand how it impacts
the design directly.
Right?
You change the design because of changingstress levels.
Mhmm.
But what about brass tacks?
(14:57):
Right?
Where someone's paying dollars for this.
Right?
Where does it help in terms of value for, let'ssay, the project itself?
Yeah.
Yeah.
It's a great question.
Well, for one, the platform itself doesn't costa whole lot of money to to spin up a study.
So we tried to make it as approachable andeasy.
(15:18):
You can download the app right now and justlook at your own personal stress information
for free.
But for the business side, there's a lot ofmoney when you when you start to go down a path
and you have to change later in the designphases.
It impacts cost because there's othersupporting mechanisms that are in place or
getting supported like utilities are beinglaid, walls are starting to, you know, to come
(15:40):
up.
And the later that you change design, it has abig cost implication.
So this is one of those tools, and there's alot of other innovative things we're doing.
I don't wanna spend all the time on EMPATH.
I know it's awesome.
But but we're gonna make more decisions earlierin the design process.
And if we can do that with this the finger onthe pulse of how other environments that might
(16:01):
relate to this current place we're designing,how they're responding to them, or is it
positive or negative, and what might work in adifferent context, that's that's gold.
I mean, that that's way better because we getto a better design earlier using historic data
or using current data about a place.
And, again, there there's other new thingsgoing on with impact in the empath world that
(16:23):
I've gone even further upstream.
So brass tacks, not everyone wants to share,you know, data about about where they're at.
Not everyone can afford an Apple Watch or notnot even afford a $20 Bluetooth band on Amazon.
So there are equity issues, absolutely,centered around a platform like this.
And it I'm not gonna mince words.
It's surveillance capital platform.
(16:45):
Now I have tried to do what's right with thething.
I can't look up, Will, in my my platform.
You're this you know, you're disassociated fromyour data.
I can ask you some questions at the beginningof the study, but outside of that, you own the
data.
You can revoke access to everything, and itwipes your data from the system.
So from user privacy standpoint, that's onething.
(17:05):
But from the business side, I think it makes alot of sense.
We have to better understand the people thatwe're designing for so we can design the right
thing.
That makes a lot of sense.
Kudos to you for thinking of the privacy side.
So for all our privacy experts out there.
And we, as Spot Migration, definitely approveof at least the concept around how you're
(17:29):
looking at data, very of you.
Well, I that GDPR ish, you you mentioned that.
I actually work with Amsterdam.
I hope I said the g right to please my Dutchfriends.
With the GDPR expert over there because one ofthe early explorations we did was back in the
city of Amsterdam, which is which is just afun, like, reconnection.
(17:50):
So, yeah, we we went into it, and and theychallenged us pretty pretty heftily about how
we approach data privacy and and disassociateit from the person that's sourcing.
So, again, when you have a cool thing, peoplepeople like to be a part of it, and they they
were really good partners early as we wereexploring this.
(18:10):
What what have you learned?
So, obviously, we talk about differentstressful internal environments.
But, like, when we think of, you know, if it'sbike lanes or if it's other transportation, you
know, ways.
What have you seen?
Obviously, you felt it right when you wereoverseas.
You you just felt like, is this so pleasantcompared to in my my home base?
(18:34):
You know, where it's like, I know the roadsbetter.
I know everything better because it's my theroute that I've, you know, made to take and all
these different things.
What is what are those big factors that do makethat difference?
That like, oh, wow.
This is these are huge aspects that we can justacross design make changes in.
There there are some commonalities like,separation between uses, green space,
(18:57):
typically, people people like.
There are trade offs that's costly, you know,to do some of those things.
But, it's weird because it's very it's verylocale based.
Like, if you're in Arizona
Mhmm.
You we can't put the same things in Arizonathat we do in in Pennsylvania.
Right?
They they won't grow.
And so it's very locale based.
There's there's different tolerances of stressbased on your locale as well as the background,
(19:20):
what you've been exposed to up to the pointthat you're collecting data.
So that part has been eye opening for me.
But, yeah, there's some commonalities that thatthat track, but then you can optimize it based
on the locale and the expectations of thosepeople in a specific area.
That's cool.
And one other thing I thought that was reallyinteresting that I learned from it, let's say
(19:40):
so I I had this happen to me.
I got hit, by a car.
Every time I passed by that, I didn't have anyproblems for for years.
Got hit, and and the woman could not have beenkinder.
It was it was a silly mistake.
Every time I passed by that spot for up alittle bit over a year, there was remnant
stress.
So that started opening my eyes to meaning,every time I passed by there, even when I by
(20:05):
myself, there was no one around me.
I still remember that experience, and it stillcaused me stress.
And the the the thing about that is when you'reoperating in an environment that you are
stressed out, you are more likely to makemistakes.
And so that remnant effect of stress waspowerful because what happens if I don't do a
good job or, you know, as an architect, wedesign a place that that does not resonate with
(20:29):
someone in a positive way.
They might start avoiding there's avoidancebehavior that that that comes with that initial
experience.
And so those those are things that now we'rewe're starting to talk about and think about as
designers.
You know, we we, again, going back to Will'sinitial question, what's the value?
Well, if we know that people are not gonnarespond well to something and we can verify
(20:52):
that in a simulated environment using, youknow, goggles and, and some some test data from
stress, Let's do that before we build anything.
So we can go we can again, there's so manydifferent directions we can go with it, but
it's just there's lots of little neat tidbitsthat have been tied to just that general
exploration of of people and spaces and places.
(21:16):
Because then you could also take it a stepfurther.
You mentioned the nurses.
Like, you could say, oh, and I'm not suggestingthat we should, you know, be mining everyone's
data if that's inappropriate or not.
But in regards to, hey.
You're okay with this.
This is for your betterment.
Awesome.
You are stressed to this level.
Therefore, we have, like, a a chill out room ora room for you to go kinda reset in.
(21:40):
And, we want you to do that or construction.
Right?
So, obviously, a construction site,extraordinarily stressful.
You have hit a level of stress that it actuallymakes sense for you to just step off the job
site for a second, take a minute Sorry.
Get get your composure, get back, you know, toeven or level, and then, you know, go back to
work.
And we're if using something like this allows,you know, alerting and things like that to
(22:02):
happen, then that's amazing.
Yeah.
So we we're still doing new explorations.
Actually, we call our nurse stations the chillout rooms trademarked
by us.
Nailed it.
But what were things like that again?
Business, you know, business dollars.
If you can save a nurse from stressing out andleaving, well, costs about $250,000 if I'm an
(22:25):
operator for of a hospital to replace thatperson and get them geared up to a point where
they're they're, you know, adding value as partof the staff, as part of the staff.
So there's all kinds of other things that,unrelated to design, but that that clients may
care about that I think some of this datastarts to to shed light on.
It's cool.
(22:46):
Thanks.
That's a what We got so many other cool stuff.
I was saying, so what else are you innovating?
You you know, that's if that's not enough.
So one cool thing I'll tease, and we're we'regonna announce this, via LinkedIn.
You'll see it, here in a little bit.
But, part of the part of the process gettingdesign off, you know, off the ground when our
(23:06):
clients have an idea, they talk to their teamsabout something that they may wanna design, and
then by the time that we wrap our heads aroundthe scope and go back and forth with them, it
can be a it can be a pretty not cumbersome, butit can be a challenge sometimes.
Right?
Because they have a preconceived notion, theyhave aspirations for a place that they wanna
create, and we have an interpretation based ona phone call or an email.
(23:29):
So we're breaking down the barriers and andcreating we've got a a gaming engine platform
that we've developed called Daedle.
It's my team loves Greek mythology, and so allof our when they're in development, we've
always picked Greek mythology kind of thingsfor, you know, to call them.
Daedal is after Daedal is the architect of thelabyrinth where there's all these weird pieces
(23:51):
that can fit together in lots of differentways.
So we're we're gamifying the early designphases of of how we scope projects and letting
our letting our clients come in and andcocreate with us.
So, for instance, let's say you're designing anew, a new industrial facility.
Right?
You're you're making batteries.
(24:12):
So you have a bunch of raw material hitting asite.
How much?
What's the volume of that?
Okay.
Well, the next step is the cathode and anodeplants.
You know, they they split up the raw materials,and they do two different things.
Well, how how much of that can these places howmuch raw material can these all go through a
process and turn it into something else?
What's what's the the cadence of that?
(24:32):
What's the volume?
What's the demands on the energy grid?
All of those things, we start to tack on layersof information.
And it's doing a simulation.
It's showing a simulation of those things.
And we have historic data about a lot ofprojects.
Right?
Because we've been in business, you know, forfor a long time.
We start to layer on.
Well, the last time we did an anode and cathodeplant, it it took this much area.
(24:53):
So now we're starting to rough out the spacethat supports a function that's defined by a
process that our clients are putting in withthe game engine.
And so, again, if you think if you're if you'redorky too, you've seen things like Factorio or
Satisfactory.
We're we're building things like that thatallow us to extract better information about
the processes, that allows us to better designthe places that support those processes.
(25:19):
So that's that's another thing, and and we'rewe're going one farther.
We've got, we're calling it a fidgetalexperience.
So you can play the game.
You can do it online, you know, or whatever,but you can also be moving blocks around that
are designated with these RU codes.
They're like a simplified QR codes withcomputer vision, with projection mapping where
you're moving a loading dock around, and itshows you on a real time adjacency matrix.
(25:44):
Hey.
Your loading dock is too close to the office.
Here's some negatives that you get from airquality as a resultant of that move.
So why don't you move back away?
And so the projection mapping is is making itred.
Like, there's a violation here that you maywanna consider.
So that's that's a cool thing.
I'm I'm really excited.
Super cool.
Oh my gosh.
We hadn't talked about that, but that's anotherthing, my team's been been really working on
(26:09):
hard, and it's it's looking sharp.
It's looking really good.
So that's another thing.
That that, again, that's super secret.
You guys get a teaser.
That's the first time I think I've talked aboutit externally, so you guys you guys get it
first.
But that's Oh,
there's here first.
Well but so let's do that.
Let's do that first with our clients.
You want me to tease more stuff just for fun?
(26:30):
Sure.
Yeah.
I think our listeners will gobble it up.
Well, it it stopped me because I I mean, I I'mseriously I can go because I love this stuff.
So that's step one, like understanding theprocesses and understanding the space that our
clients expect us.
Well, what's next?
We have a massing.
We have a rough area, you know, of spaces nowthat we have to accommodate, and we might even
(26:51):
have some rough layout of this the space.
Well, we're doing a real time rendering engineon top of it.
So as the the digital cameras are picking upthe layouts or we're having the the digital
pass off from our game engine, we have, what wecall Gia Imagine.
Gia is our internal chatbot.
She she is it's it's we've engendered her.
(27:12):
She's female because she solves all ourproblems.
But we have modules for Gia, and it stands forgenuine ingenuity assistant.
Genuine ingenuity is the Gresham Smith tagline.
Harkens back to the early you know, what I wastalking about.
It's our a little little aside.
Genuine means we care about we care about ourwork.
We care about the impact, so we do it in moreingenious ways.
(27:34):
We we tackle things in in kind of moreintelligent, more smart ways.
Gia imagine starts to pick up thisinterpretation of amassing and do real time
verbal input.
So we can say, hey.
Here's a reference image.
I really like the glass on this panel, but Iwant it to be darker blue, whatever.
Gia imagine is taking that and doing a realtime rendering at the same time.
(27:56):
And that's the mid that's the middle piece.
That's working too.
So the last piece of that is another iterationof Empath.
It's called Empath React.
Just like I'm I'm seeing your faces.
Again, keeping the design process very human.
We're good.
You know, I'm seeing you both smile.
That's great.
That makes me feel good, but I'm doing thatbecause of a combination.
I'm tracking I know your mouths are a littlebit agape.
(28:18):
I see your teeth.
I see you have a a slight smile.
I know you're squinting, so you're interested.
Those are all things that we do every day.
Empath is is something else that that I'vebuilt with with my team over the past cup maybe
three or four months.
It's been fast, but it's an emotional model.
So it looks for known points on your face.
It looks at the math between those known pointpoints and the angles that they change, and and
(28:43):
it picks up, are you responding well to whatI'm saying or showing you?
And so that's the the last layer of this is andfor it does verbal sentiment analysis through
in real time too.
So as I put down a selection of lightinglighting or furniture in front of you, You
could say, oh, I love the brown couch.
I hate that light.
(29:03):
That would go good with this.
It's it's listening to you, and we're passingthat back to g imagine.
So it's swapping things out based on yourreaction.
Oh.
And The design elements.
Yeah.
So that Are
you talking about, like, even, like, microfacial expressions?
Yes.
It's picking that up?
Exactly.
That's exactly Oh, jeez.
Okay.
That is wow.
(29:25):
That that's impressive.
Well, the now the tie back is you you justheard, like, the beginning part of we wanna
design better places for people.
We wanna make it fun for our clients, and wewanna get better information out.
We wanna show them how we augment, how we howwe're not just replacing our our professionals
because these these are the the people that Iwork with.
(29:47):
You know, we have 1,200 employees at GreshamSmith.
They're brilliant.
I don't wanna replace, you know, their talent.
So let's augment.
So we're we're making the the creative partfaster, and then we're using again this this
next iteration of empath, empath react to pickup, you know, slight variations and changes in
facial cues and verbal sentiment analysis todesign the right stuff faster.
(30:10):
And so we're we're lumping all this togetherinto the Gresham Smith design the Gresham Smith
design lab.
So So that that's close.
If you so if the term hasn't been coined yet,you heard it here first.
It's not vibe coding.
It's vibe architecting.
Oh.
I'm gonna submit for a trademark, and I'll I'llgive you credit, Will.
(30:32):
Appreciate that.
That's just a glimpse, and I didn't wanna spendtoo much time on just the impact because it
sparked that little bit of a challenge, likehow we're approaching design.
We could be doing that daily.
We could be evolving and making this process somuch more fun and so much more enjoyable for
for 1,200, you know, Gresham Smith architectsand engineers and program managers.
(30:57):
Let's challenge all of that.
And and, you know, Will, I know we talkedbefore during introductions before about
digital transformation.
I hope now you see this is what I'm talkingabout.
We do all of these things manually.
Now we have tools that we've either developedor that we, you know, can guide to to really
evolve what Gresham Smith is and and define apath for us that that is the future of design.
(31:21):
And I think we're we're on to something.
Is automation part of your scope as far asinnovation goes?
It is.
Yeah.
It absolutely is.
So then let's talk a little bit for a second.
Mhmm.
What's the secret to automating things?
(31:41):
So it's it's really it starts with the team.
I don't know the so let's let's take forinstance, a mechanical layout of a plant.
Well, our practitioners know how that goes.
They can do the manual process all day long andin their sleep almost.
Me as my background, you know, as as Justinsaid, his background is transportation
(32:03):
engineering with this weird overlay oftechnology.
I will never learn more about that process thanwhat they already do.
So it really starts with the team.
They have to identify things that they wouldlike to target for automation, and then we are
very reliant on them to spend the time tomanually map all the processes associated with
that manual working through it.
(32:24):
Woah.
Woah.
Woah.
Hold on.
Hold on.
You said you rely on them spending time.
That's right.
By any chance, is he, like, your middle namechief cat herder also?
It it can't can't feel like that, but you wouldyou would be surprised, when they get it,
they're really willing participants.
So it's less herding cats and moreorchestrating slight chaos at the beginning.
(32:50):
K.
That actually talk or speaks to the culture.
Right?
Because, like, it is the idea that we are goingto, create better things.
Right?
Like that, you know, and however you chalkchalk that up.
And that's not just the innovation team.
That is across the board.
We are gonna create a better client experiencedoing that.
So if we automate things that will create thatbingo bango, here it is, which, is a beautiful
(33:14):
thing to see when it comes to a company of yoursize to have culture embedded that deep.
I I love that you brought this up.
Our our CEO, Rodney Chester, has been reallyinstrumental in in helping to guide with all
the things that we're talking about today.
He set this up for success, and you can see itwhen you walk through the on people's faces.
They're excited.
And Rodney, again, I I don't wanna I don't wantthem to think I'm bragging on them too much on
(33:37):
the on the podcast, but I'm pretty sure.
But but he's come out and said, I wanna makesure that we have the best culture of any AE
firm out there.
And then he turned to me and said, we have tomake sure that when people choose to spend
their time at Gresham Smith, that they have afull long lasting career that they feel fully,
like, invested and they're getting somethingout of it too.
(34:00):
Well, to me, I can't turn on anything that'sgonna short circuit that.
And so going back to automation, the reason Istart with the team is I really want them to be
guiding and leading and picking the rightthings to automate.
There we could probably automate the world.
Like, there's so many things we could beautomating, but it short circuits.
It either creates knowledge gaps.
It creates, you know, issues, for forbillability on things.
(34:26):
There's all kinds of other things that we haveto be cognizant of, and it there's trade offs
in all of this.
And the other thing too, we're we're, you know,we're stamping things.
Architects and engineers certify our work, andand I'm very cautious of automating something
that could be interpreted as, skipping thestandard of care.
That's that's a nonnegotiable for us.
(34:46):
So it's all of those things yield somethingvery positive.
But, yeah, it starts with the culture.
I love that you brought that up, Justin, andthat's something that's really near and dear to
to our our executive team at Gresham Smith.
So to go, you know, a step a step further intothe automation side
Yeah.
(35:06):
How is it affecting work?
And you said there there's some you described aphrase, and I want you to kind of maybe either
zoom out or zoom in.
I'm not sure I'm not sure which direction youwanna take this.
But when you talked about sort of the care andmaking sure that there's certain pieces that
(35:31):
are not skipped, are you saying that there arecertain pieces that you could automate, but
you're, I guess, making a conscious decision ofnot automating?
Absolutely.
Yeah.
Yeah.
Talk about that a little bit.
Yeah.
So, especially when you have, like, onboarding,you bring a new engineer or architect on.
You want them to learn the actual steps.
(35:52):
You want them to learn the process for how tothe Gresham Smith way that we approach design.
If we skipped those things, if we automatedthat early experience, you're in effect
undercutting their ability to grow as aprofessional.
You have to be we have to be very careful aboutthat.
So, yeah, we're not we could be automating alot more that we're we're purposely not because
(36:13):
we wanna build great designers, architects andengineers, and and everything else.
Consultancy services we do.
We don't wanna short circuit their ability tofeel fulfilled, to grow as an individual and
professional, and and understand the decisionsthat they're making on a project have real
implications.
So, yes, we're not we're not automatingeverything, but we do choose to automate some
(36:35):
things that are highly tedious that drivepeople insane.
Absolutely.
But we still walk them through what it's doingtoo just so they're they're aware.
Are you able to share, any of these, we'll calltedious things?
So just to give an idea to people that are outthere, like, what you think is worth automating
(36:56):
and spending the time?
Because it takes time to automate.
Right?
Like, automation is great, but there's a lot oftime that's spent on doing the automating
before you actually get reap the rewards.
Yeah.
So so there's several.
One thing in particular that I've been reallyproud of is there we have some really senior
level folks that have built these massprocesses in Excel to do calculations and
(37:19):
checks on things like electrical layout andsupportive systems.
And the neat thing about this is it's it takesa lot of time.
You get a you get a a plant layout.
Again, we'll go back to industrial sites thatare very complex.
They have lots of different manufacturers thatthat you have to support things moving in that
that do things with raw material or or part ofa process.
(37:40):
Just the electrical side of that.
Let's say one of the equipment manufacturersswaps out something.
That changes everything.
It changes the the entire electric grid and allof the panels and all the support mechanisms
behind that.
Every time one of those change, and and thatchanges a lot.
So you take a junior engineer that is manuallygoing through the same process thousands of
(38:04):
times, and you automate it and say, hey.
We just saw you swapped out this.
We can read because it's a smart object.
What the load is, we're gonna go all the wayupstream and change all of the panel
configurations, And here's the output, andhere's a report of what was changed and why it
was changed.
Check check this.
Do you agree with it?
Yes?
Okay.
Well, you just saved literally two thousandhours.
(38:28):
Oh, jeez.
And I'm not
I'm not I'm not exaggerating at all.
That that's that's exactly how that processwent.
And I will say kudos to to that team.
They invested about a month's worth of time onthe front end to to convert that Excel into a
logic map for us to follow.
(38:48):
And then my team was able to automate it inabout two days.
Okay.
Well, that's worth it.
I mean, so took a month.
So two thousand hours, that's basically ayear's worth of man hours.
A two thousand hours on a on one project whenone thing swap
out.
One thing changes.
(39:09):
Oh, jeez.
Can be the scale of it is in enormous,enormous, and it's tedious work.
Why would we you know, we want people to beexposed to the work, but we don't wanna bore
them to death.
You know, we don't want to get them superexcited to come to work for an architecture
engineering firm and then spend all dayreplacing, a decimal place in BIM.
(39:32):
So but the other thing too, it's not just timesaved.
It's it's better for risk management.
We've we've eliminated manual everything wherewe can on things like that.
So the me getting tired after six hours ofdoing what I just described, maybe I do miss a
miss a number.
Maybe I put a nine instead of a six.
(39:52):
Well, because it's automated and because weverified the automation process, we we get
better delivery that's less risky.
Yeah.
Removing the human error, essentially.
That's it.
Yeah.
No, that's and morale.
Like, yes, x amount of time.
Sure.
But like when you're doing something that suckslike that and you're doing it for that long,
(40:14):
it's like, oh yeah, that's such adeterioration.
And then you're unhappy.
And then if you're unhappy, those around youfeel bad, and it goes you know, it chunks on
and on and on.
I don't know how many modern interns would bewilling to move a decimal point, and spend two
thousand hours, doing that, across an entirebuild.
Yeah.
You're you're absolutely right, Will.
And and when you think about, okay, an internthat has two choices, they've got they can come
(40:37):
to Aggression Smith and have all this coolstuff and support and be a part of, you know,
evolving, and and we we let them.
We have junior level folks on our futures labis what we call it.
They give us ideas and we we evaluate those.
They can be do that or they can go and do allof these manual processes at another firm.
Well, guess who's gonna win?
And there's a war for talent like that's a realthing.
(40:58):
Finding enough folks that really wanna do whatI'm describing to you and have have a positive
impact on design, it's still tough to get them.
So yeah.
What has the, essentially the investment intechnology automation, whether it new new
programs, new innovations, or just thesenondesimal moving automations.
(41:21):
Like, what has been that impact when it comesto hiring and retention?
Well, we're we're growing.
I mean, we're that's the thing.
We have, I think, a 90 something.
I can't remember and I don't wanna quote, butit's north of 90.
It's way up there for employee satisfaction.
Our retention rates are very high.
We have very low turnover.
And typically, when we get to offer stages, wehave a lot of success, not just for innovation,
(41:45):
but just in general, the overall culture, howwe approach approach design, and the accolades
we get for doing good design.
So it's it's a innovation's a piece.
It's not the answer.
It's a piece of the overall value that that afirm like Rushing Smith can bring.
But, yeah, it it works.
You're you're so humble, Mike.
You should be like, it's all innovation.
(42:06):
That's the reason we we exist.
It's not.
And I think if if I was on here saying that,you should not be talking to because
I can care.
I'm joking.
Yeah.
No.
No.
No.
I get it.
But, yeah, it is a it is a piece of it.
But there's some tried and true things that arealways gonna exist in design that we don't
wanna disrupt.
But there are other things right fordisruption.
Well,
(42:29):
we don't wanna, you know, we don't wanna diginto any future innovations and, you know,
however, are all innovations is it seeneventually at the client end?
Would you would you say that the client sees itor customer, however you say that?
(42:50):
Historically, we've done a a bad job of that.
I I'd like to sit here and say everything'sbeen smooth sailing.
It's not.
We're we're rebuilding this plane as we fly it,as some would say.
We're figuring a lot out as we go.
And I would say, historically, we've done a abad job of that.
Now, though, we're taking a different tackwhere we are since we've kind of established
(43:12):
the space of trust with our internal marketsand corporate services inside of innovation,
our markets are starting to say, hey.
We're gonna have clients come and evaluatethis.
Is that cool?
Yeah.
That's great.
And our clients may say, man, I wish you woulddo this and this, but we can do that.
But having them come to the table and and Ithink us being willing to let our guard down a
(43:33):
little bit and and I like the
word
cocreation.
My my team will make fun of me because Iprobably overuse it to them, but I think it's
important.
And we can do that not just with design.
We can do that with these platforms.
And so it's been fun to to tease these out.
Like, all the things that I've just describedto you, clients have seen, And and they're
helping to shape, and and they're gonna be partof the early adopters.
(43:56):
But but we're building it kind of together.
And so, yeah, we've gotten a lot better at atbeing willing to to pull the veil back a little
bit, let let our clients see rough ideas beforethey become, like, tangible.
So, yeah, I I think that's gonna continue.
And and I think there's gonna be a lot more ofthat that model moving forward.
(44:17):
You spoke about Gia as sort of your internal AIchat.
Mhmm.
I'm assuming you were part of that process.
Yeah.
For companies that haven't actually dug in.
Mhmm.
Or they're at the very beginning stage of theirprocess around choosing the direction and
(44:44):
policies and all that.
Mhmm.
What are some learning lessons that you guyshave learned Mhmm.
And maybe if you'd had to do it all over againor would have just either been better if you
sped things up.
Right?
What are some of those learning lessons thatyou learned as you were implementing
Sure.
Well AI.
(45:04):
Yeah.
It's a great question.
And we were we were really early adopters.
Again, I'm fortunate to have, you know, anexecutive team and a board that that supports
explore or exploration early.
The things that weren't well was we went out ofthe the gate with, we had a governance policy
that we we drafted almost immediately.
We had a usage policy as we start to adopt newtools, even before we adopted them.
(45:27):
As we start to adopt new tools, here's what weexpect you as a user to kind of concentrate on,
which is not skipping the standard of care, notobviously using the tech for bad stuff, not not
skipping parts of the process.
Then we started digging into specific tools wewanted to use.
The first version that we unveiled of Jia wasresided in Teams.
(45:48):
This was two yeah.
Two two years, almost three years ago, so itwas early.
We put an early AI agent in Teams.
We coded it ourselves, and it was okay.
You know, it kind of unveiled that, hey.
We're we've adopted something early, but it hada lot of hiccups.
Like, the history was funky.
(46:10):
You couldn't reference things without manuallygoing back and doing kind of the the now the
current version of of most platforms.
You go back and forth with chat.
The so that was a lesson learned, but that wasthe option.
The next lesson that I would say is really diginto the buy versus build.
So for Gia, which is an internal platform thatwe've coded, it actually has more functionality
(46:31):
than OpenAI's offering.
We built it for about $12,000 total, and ourmonthly reoccurring cost is between 5 and $600
on API usage.
APIs, we did that specifically because the datathe data privacy inside of an API connection is
is secure.
It's protected.
(46:51):
So if we turn the same thing on using aprebuilt system like an OpenAI or Microsoft
Copilot, which, you know, there there arepluses and minuses to that, that would have
costed anywhere between 250 and $450,000.
But we have expertise.
We have the expertise internally, and so webuilt something that's a really great use of
(47:13):
dollars.
That first chatbot, the first the v one, whatwas that to the did the whole company get
access to that, or was that just innovation?
Okay.
So whole company did.
What then was the use case for?
Like, what what were or what were people usingit for?
Or what people what did people think they coulduse it for?
(47:33):
Maybe some Yeah.
Question.
Pretty much anything.
I mean, it was all over the place, and we werepretty open and honest.
That's another thing, advice I would give.
Lesson learned is be open and honest that, hey.
We're watching.
Don't put Yeah.
Terrible stuff in this.
You know?
And and we trust our folks.
That that didn't happen.
But we told people, hey.
We're watching this, and we did it with youknow, what we wanted to do was if we find a
(47:58):
bunch of people using it for design processesor challenges.
Mhmm.
Let's invest more of our time from theinnovation side in IT and MIS on focusing on
those types of of agentic systems or the customknowledge bases.
If people were using it for recipes, well,we're probably not gonna invest a lot of time,
and there are people that use it for recipes,obviously.
But but we tracked it to make sure that we'reinvesting our time in the right types of
(48:22):
knowledge bases or or add ons for our internaloffering.
And we launched Gia chat version two, and we'vegot generative image and video and everything
now too.
All all Okay.
In in house.
That helped to guide, you know, the types ofthings and system we wanted to build.
And so we do have all of those custom knowledgebases.
(48:42):
We've got, automation scripts that go back andforth.
We've got NCP connections now inside of of Giachat that connect to other things.
I can't talk too much about it.
There's a lot of there's a lot of secret stuffaround that, but it's doing a lot more for us
because we were open and honest with the peoplethat we're engaging with the first version.
So that that, I think, is another lesson there.
(49:04):
And the last one, lesson learned going back tothat question, I know you've asked other
things, but this is an important one.
Just because you turn on something or youinvest in something and you expect everyone
else to be just as excited as as, you know,maybe maybe us on the phone.
That just doesn't happen.
And so we we started coming out of the gatewith, here's use cases, and we let those people
(49:25):
that develop the use cases own them andrepresent them to the firm.
We have we have, innovation host quarterly,like, firm wide calls where we showcase a bunch
of cool stuff happening.
Again, trying to get more people excited, butthe use cases and then the training videos is
you gotta do that because not everyone is goingto be as curious or explore the same way that
(49:48):
that I do.
So, yeah, that's that's
Does g o two now you just prompt g o two tomake those videos for you?
Because it's just like that.
So we we have started building commercials.
You'll see all the things I teased earlier,like the the big new we've got commercials that
we've developed.
Yeah.
They've got voiceovers and look like we'vehired a production company.
(50:08):
Yes.
Nice.
Yeah.
Also, looking for recipes could be a goodthing.
There could be a company cookbook.
That that's like a document that could live onyour SharePoint.
I I love I love,
you know, cooking and and things like that.
And and ultimately, whatever.
You know?
I'd rather somebody be curious about it andtinker with it than oblivious to it.
So
(50:29):
I also think that in, how how it's set up, butlike getting them to just use it, right?
That's the biggest hurdle because then onceonce, and this is what certainly we've seen
when somebody's interested about AI, it's like,well, are you, what are you doing with this?
So I've not done anything.
It's like, okay, Well, just start playing withthe chatbot.
(50:49):
Right?
Like, just start start there.
And then what will happen?
You'll start getting a lot more ideas of like,oh, this could work for the thing I'm dealing
with in business.
Like, oh, but like this totally I could usethis this way.
And it's like, yes.
Exactly.
And that's how innovation works.
Right?
Like, that's the just explore the next thing.
It's it's fun.
(51:09):
Like, last week, we gave a presentation,current use cases to our owner's group, and you
saw light bulbs just going off across the room.
They're like, oh my gosh.
I should be doing this.
It would save me so much time or I could beputting together such a better proposal
response.
And so that just level of exposure to somebodydoing something really unique and saving a
(51:30):
bunch of time or producing something betterwith the help of this augmenting tools.
It's just making us better.
Yeah.
Light bulbs across the room.
It was great.
That's awesome.
Yeah.
Justin, I think that we are at a good point togo to our last question because otherwise, we
(51:52):
are gonna go on like Joe Rogan for four morehours.
I don't know.
I have it in me.
Yeah.
I don't yeah.
You're, East Coast time.
Yeah.
Awesome.
So, Mike, we love to ask this question toeverybody.
If you could go back twenty years, what advicewould you give yourself?
(52:12):
Oh, that's a really good question.
What advice would I give myself twenty yearsago?
I would say, as with anything, I struggled, youknow, when when you're intelligent, when you
when you can process lots of things reallyfast, you have lots of big swings.
You got lots of dips along with the peaks.
(52:33):
I've talked mostly about peaks today, and Iwould say it's worth it.
Like, stay curious.
Keep digging.
All of those little moments that I spentteaching myself Pascal and something that I
didn't think would amount to anything, but Iwas just curious about, they all add up into
something really cool as long as you pursue it.
(52:54):
And I think that marrying up something that Ilove to do, which was I wanted to influence
places, and I've married it up with that lovefor technology.
Pursue that passion.
That that that the only reason I'm able to talkto folks like you and actually feel good and
confident about the the future of my of thiscompany, Gresham Smith, is because of that.
It's because I've I have been very fortunatethat I've been in a position where I can marry
(53:18):
up multiple different passions and and helphelp help a lot of folks.
And I would say, I didn't mention this before,but it works.
We just won, Fast Company's one of the mostinnovative companies in the world.
That's prestigious.
Well, it's but it works.
And so it's, again, it's just that that staycurious, marrying up the passions, and pursuing
(53:41):
it.
It it it ends well.
So, yeah, I'd say that.
Last words.
That's that's This
has been a ton of fun.
If people wanted to get ahold of you, what'sthe best way for them to do that, Mike?
I would I would direct them toaggressionsmith.com or or or find me on
LinkedIn.
I'm pretty active participant on on that.
So, yeah, happy to engage with folks.
(54:02):
I'd love to hear their thoughts about some ofthe new innovations coming from aggression
smith.
Awesome.
Awesome.
Awesome.
I'll make sure I put, your LinkedIn, and thewebsite in the show notes.
And is there anything else you'd like to tellthe people before we say goodbye?
I would say you're getting ready to see so muchchange.
It's gonna be a really exciting time for notjust innovators, but just I I I listen to your
(54:25):
podcast.
I love what you all do, and there's a lot offolks that really care about about all of these
things.
And I think you all do a great job, so there'sgonna be so much more opportunity to capture
this and to do cooler and more impactfulthings.
Just get ready.
It's just it's gonna be a lot of fun.
Awesome.
That's awesome.
Thank you.
Thank you for listening, Mike.
(54:45):
We appreciate it.
I appreciate the opportunity.
Thank you guys so much for hosting.
Awesome.
Well, listeners, I hope you had as good of atime, and maybe mind was blown with, the
innovations that are coming out of GreshamSmith, as me and Will are.
But until next time, adios.
Adios.
You guys.
Thanks for listening to building scale.
To help us reach even more people, please sharethis episode with a friend, colleague, or on
(55:11):
social media.
Remember, the three pillars of scaling abusiness are people, process, and technology.
And our mission is to help
the AEC industry protect itself by makingtechnology easy.
So if you think your company's technologypillar could use some improvement, book a call
with us to see how we can help maximize your ITcybersecurity strategy.
(55:35):
Just go to buildingscale.net/help.
And until next time.
Keep building scale.
I've got a really nice route.
(00:01):
I looked compared started comparing ridesbetween The Netherlands and The United States,
and I had way more spikiness in my heart rate.
And there was elevated levels of heart ratewhere it shouldn't be.
Like, it was a very comfortable ride.
And I was like, there's something behind thisdata.
And that that was what sparked, hey.
I could probably use this to do a calculationand then have people share their data and and
(00:23):
target places that need to be improved.
So that that's the origin story, and I I lovethat.
Have you ever wondered how successfularchitecture, engineering, and construction
companies scale their business?
Or have you ever wanted guidance on how to getmore growth, wealth, and freedom from your AEC
company?
Well, then you're in luck.
(00:45):
Hi.
I'm Will Forrette.
And I'm Justin Nagel, and we're your podcasthosts.
We interview successful AEC business leaders tolearn how they use people, process, and
technology to scale their businesses.
So sit back and get ready to learn from theindustry's best.
This is
building scale.
(01:07):
Hey, listeners.
It's Will here.
Our mission is to help the AEC industry protectitself by making technology easy.
If you've ever listened to our show, then youknow that the three pillars of scaling a
business are people, process, and technology.
So if you suspect technology is your weak link,then book a call with us to see where we can
help maximize your company's IT cybersecuritystrategy.
(01:30):
Just go to buildingscale.net/health.
Today's guest is Mike Sewell.
Mike is the director of innovation and a boardmember at Gresham Smith, one of the nation's
leading architecture, engineering, andconsultancy firms.
With over two decades of experience, Mike isknown for pioneering safety focused design and
leveraging emerging technologies to improveinfrastructure for vulnerable road users.
(01:55):
He's the creator of the patented empathplatform, the first of its kind to quantify
emotional response in different environments.
Empath earned honors, major honors, includingFast Company's world changing ideas award and
architect magazine's r and d award.
Mike also serves on the board of the league ofAmerican bicyclists and has testified before
(02:15):
congress on multi module safety andinfrastructure funding.
His work at Gresham Smith continues to blendtraditional engineering with cutting edge
innovation, all aimed at creating healthier,safer, and more connected communities.
And with all that said, Mike, welcome to theshow.
Will and Justin, I really appreciate you,affording me the opportunity to come and speak
(02:37):
about something I love, which is this weirdkind of mashup of design and technology.
So I'm I'm really thrilled to be here.
No.
It's super fun.
I I get giddy, thinking about what we're gonnatalk about because the empath thing is is wild,
and what you're doing as an innovation leaderis just fun.
It's just totally down our our pipe.
(02:58):
And it totally tickles the funny bone when youthink about, like, you know, there's a lot of
talk about AI, but this this is this isinsanity.
This is a place that I did not fathom that itwould go, and yet here we are.
So don't wanna keep you waiting.
So Justin, why don't you why don't you get usstarted?
(03:18):
Yeah.
So, Mike, I said all these wonderful nicethings, but tell us your origin story.
How did you get in engineering?
How did you eventually get into the innovationpiece of that?
And, tell us a little bit more about GreshamSmith.
Yeah.
Sure.
Well, in the intro, I might need to prop up myhead for a second.
That was very kind of you.
Thank you for going through all the accolades.
But, honestly, it's it's been a labor of love.
(03:38):
Started with with a background in engineering.
I real I chose engineering because, you know,when I was growing up, we were very reliant on
a lot of public public outreach, public thingsthat were not in place that really made our
jobs a lot a lot more difficult than I thought.
If I chose engineering, I could I couldpotentially help the communities that I love,
and and prop up other folks.
(03:58):
So that's the reason I gravitated towardsengineering, which has been a really fulfilling
career in and of itself.
But then the tech side, I'm a dork.
You'll figure that out over the course of thispodcast.
We're pretty pretty you know, you'll figurethat out pretty quick.
But, I've always been just enthralled withtechnology.
So going back as far as, the mid eighties whenI got my first home computer that was, you
(04:21):
know, gifted to the family, I I taught myselfeverything, autodidact all the way.
Like, never have taken any, computer courses,just have poured myself into finding out and
staying curious about it.
And so I got a lot of background with that too,but then I had the opportunity, at at Gresham
Smith to to lead the innovation program, whichis kinda wrapping that technology side and that
(04:44):
love for technology around the the genuine carethat I want, as an engineer and and now
architects as well.
So it's just it's that's a little bit of backabout the background.
I'm happy to go into more detail, but that'sthat's the origin story.
Alright.
Well, let's hit on that core purpose of, youknow, you essentially wanting to make your
community better or your communities better.
(05:04):
Right.
What oh, let's let's dig into that a littledeeper.
What what does that truly mean to you, and andwhy does that drive you?
Yeah.
Well, it's it's pretty simple.
You know, while some of us can choose forms oftransportation, some of us choose, how we
interact with an environment around us, a lotof people don't.
I didn't.
(05:25):
When I you know, when we were growing up, we wehad it rough.
And, you know, now, you know, you you talked alittle bit about my, background with and love
for multimodal kind of engineering andtransportation.
A lot of people, that's not a choice.
Like, that's that's the only viable solutionthat they can afford to get around.
And I think that really solidified.
(05:45):
Those folks don't have a voice, those folks arevery dependent on places that I influence
potentially as an engineer or designer ofplaces, so why not really invest myself in
doing what's right by by the communities thatdepend on better and and really should be
demanding better places and better connectionsto those places.
So it it's as simple as that.
(06:06):
Just looking out for people that depend onplaces and infrastructure.
Alright.
I'm not gonna try to hide the headline here orbury the headline as they say.
Let's dive into the empath stuff.
So what Oh, yeah.
First, what is it?
And second, where did it start?
Where did this idea come from?
Sure.
Yeah.
So in innovation, this kinda kicked off empathat at Gresham Smith.
(06:30):
This was the impetus behind my current role,which was I started really challenging,
engineers like data.
Right?
You you we like a lot of numbers.
We like a lot of supporting information, andthere was a gap.
I think in general terms, I'll pick onengineering, but but one thing in particular
always bug me about engineering is we wait alot of times to challenge the status quo until
(06:51):
something catastrophic happens.
We we wait until someone loses their life.
We wait until someone has a serious injury, andthen we're like, oh, we should probably take a
look at this roadway or we should take a lookat this space that's caused, this problem.
The problem is, like, that's real life.
That's that's a that's somebody that that hadto be on a piece of infrastructure or in a
(07:11):
place and something, unforeseen happened tothem.
That's a terrible data point to wait for.
And so that was the spark, behind Empath.
What it is is it looks for precursorinformation.
It's I will dork out for a second, but it'sit's using low polling heart rate data on any
modern wearable or Bluetooth device.
(07:32):
It is looking for nonlinear transforms of heartrate.
If you look at the two calculations primarilythat calculate stress, like if you go into a
stress test, it's it's using EKG level data,which drains batteries.
Mine uses low frequency heart rate data, so youcan record all day long everywhere you go using
using not a lot of battery, and it marries thatdata up with the location information.
(07:58):
K.
So what that in of itself, it's looking forstress.
It's it's calculating fight or flightmechanisms.
So think of it if you walked out and got buzzedby a bus on on the side of the on the side of a
road.
You're not losing your life.
It makes you very uncomfortable.
That shows up in data now.
And so comparing that to the original EKG, thetwo primary methodologies for calculating
stress, they're just they're just slopeanalysis.
(08:20):
And so I dumbed it down and created a novelcalculation for stress.
I had it vetted with Vanderbilt.
And then that that part is fun, in and ofitself, but the real power is aggregation.
So when you've got a lot of people experiencinga similar type of response, well, it's no
longer an individual user.
It's the environment causing that stress.
(08:41):
And so you start to see it's painting a picturenow where if we can go upstream from somebody
having to have a serious injury or or losetheir life and we start looking at these
leading indicators of people not feelingcomfortable in space.
Well, let's go fix that space before, you know,somebody does have a problem.
That's the background.
That's what it is, but the the origin story iseven cooler, I think.
(09:04):
Again, I'm a bike guy.
I I was invited as part of my infrastructurerole to audit, a group of people biking across
The Netherlands.
So I was in The Netherlands, biking across thecountry.
We we biked in miles.
It's about 220 miles across the com thecountry, and I was just, like, hit.
I was outside of Leiden.
(09:25):
I'll never forget it, and we were in this biglong group of people biking, and it was just
this wave of appreciation for how they puttheir infrastructure together hit me.
And I was it it was just magic.
It was beautiful.
It was intuitive.
It was comfortable.
There was plenty of accommodations.
You know, I knew exactly where I was.
Everything just lined up and checked all theboxes for places and for infrastructure.
(09:49):
And I remember I pulled over.
I took a picture of that spot, and I was like,I wanna calculate this.
And so then I, you know, I've mentioned I'm adork.
I I record everything.
I've got I've got sensors on me at all timeswhen I'm on my bike rides.
It know it knows where I'm at, and I startedlooking at the data.
When I got back to the stateside, I startedreally digging into a normal commute.
I used to bike commute every single day.
(10:10):
Just love love that mode of transportation.
And just even when you you feel like, hey.
I've got a really nice route.
I look compared started comparing rides betweenThe Netherlands and The United States, and I
had way more spikiness in my heart rate, andthere was elevated levels of heart rate where
it shouldn't be.
Like, it was a very comfortable ride.
And I was like, there's something behind thisdata, and that that was what sparked, hey.
(10:34):
I could probably use this to do a calculationand then have people share their data and and
target places that need to be improved.
So that that's the origin story, and I I lovethat.
It's just that that hit me there in such apowerful way that I was like, I I know that I
can figure this out, and I'm not, you know, I'mnot a medical.
Medical is not my background.
(10:54):
But, again, engineering is all about data.
It's all about dissecting a problem and tryingto find a new solution.
I know we're gonna talk about digitaltransformation, things like that, but that was,
like, weird, confluence of a couple differentpassions just all coming together at the right
time and in the right place.
So what, what what has come from this?
Right?
So this sounds cool.
(11:15):
Right?
Like, as a listener, it'd be like, okay.
So you understand that certain places are morestressful?
Totally down with that idea.
What has now come out of this?
Yeah.
So a a few things.
For one thought leadership.
I think the demonstration that designers atGresham Smith and I I wanna I wanna give kudos
to the company.
They invested in in me exploring this.
(11:38):
Thought leadership for number one.
We care about the places we design, so whydon't we use intuitive new ways of figuring out
how to do that even better?
That's number one.
We deployed it in places like Denver.
They have a massive population that that lovesmultimodal travel, and and they were they were
really, an early adopter.
(11:59):
So they allowed us to do a pilot test with withDottie out there.
We deployed it.
We had we collected about three and a halfmillion data points across the city.
I I can share graphics of it later, but itreally highlighted areas that they wanted to
target for improvement that there were no crashdata.
So it highlighted again, there's discrepanciesbetween the historic way that we've approached
(12:22):
projects and what the stress data can tell us,an aggregated stress model can tell us that,
hey.
There's a problem here.
Let's let's aggregate all these things.
Let's, prioritize different projects that wethink will have an impact, and we can do quick
fixes to see if that's addressing the problem.
And if it is, then we can invest more money.
So that kind of informed how council wasstarting to allocate dollars.
(12:44):
Lots of other cool things came of that project.
Since then, though, we've also started tryingto take it indoors.
Think of other high stress locations likeairports.
You know, no one thoroughly enjoys the day thatthey have to go through and work out through
security, but the next step for us would belet's use it for a reactive environment.
So let's say you are arriving at an airport andyou say, yeah.
(13:06):
I'm gonna share with a Delta via AggressionSmith my experience journey.
We can map that now.
We can we can highlight specific areas wherethere might be confusion that shows up in
stress data and target those for interventionslike, simple way finding and signage, things
like that.
We've also deployed it I know I'm go I'mrambling now, but it's so there's just so many
different directions.
(13:28):
Think of high stress environments like thehealth care industry, which we design a lot of
of hospitals.
And we're constantly changing, you know, thetypes of environments we create to influence
either, a customer experience or to supportprofessional staff that are in those
environments.
Those people are in really high stress jobs allday long.
So recently, you saw this come to a head duringCOVID.
(13:50):
We had nurses burning out.
Stress.
It's stress.
Mhmm.
So can we start to detect when elevated levelsof stress start to, you know, inject them from
their sweet spot, the the the flow state, youknow, as as people call it.
Let's let's do a mandatory respite in an anAggression Smith designed respite room and get
(14:11):
them back to an optimal level to carry out ajob.
So it started to open our eyes.
There's a lot of other ways that we can applythat kind of approach to positively influence
the design of places.
So I hear I hear the positives.
In terms of, I'm gonna put my business hat onfor a second.
(14:35):
Like a lot of dollars are invested and soundslike, you know, for someone that, you know,
when we talk about the business hat, how doesit impact either I understand how it impacts
the design directly.
Right?
You change the design because of changingstress levels.
Mhmm.
But what about brass tacks?
(14:57):
Right?
Where someone's paying dollars for this.
Right?
Where does it help in terms of value for, let'ssay, the project itself?
Yeah.
Yeah.
It's a great question.
Well, for one, the platform itself doesn't costa whole lot of money to to spin up a study.
So we tried to make it as approachable andeasy.
(15:18):
You can download the app right now and justlook at your own personal stress information
for free.
But for the business side, there's a lot ofmoney when you when you start to go down a path
and you have to change later in the designphases.
It impacts cost because there's othersupporting mechanisms that are in place or
getting supported like utilities are beinglaid, walls are starting to, you know, to come
(15:40):
up.
And the later that you change design, it has abig cost implication.
So this is one of those tools, and there's alot of other innovative things we're doing.
I don't wanna spend all the time on EMPATH.
I know it's awesome.
But but we're gonna make more decisions earlierin the design process.
And if we can do that with this the finger onthe pulse of how other environments that might
(16:01):
relate to this current place we're designing,how they're responding to them, or is it
positive or negative, and what might work in adifferent context, that's that's gold.
I mean, that that's way better because we getto a better design earlier using historic data
or using current data about a place.
And, again, there there's other new thingsgoing on with impact in the empath world that
(16:23):
I've gone even further upstream.
So brass tacks, not everyone wants to share,you know, data about about where they're at.
Not everyone can afford an Apple Watch or notnot even afford a $20 Bluetooth band on Amazon.
So there are equity issues, absolutely,centered around a platform like this.
And it I'm not gonna mince words.
It's surveillance capital platform.
(16:45):
Now I have tried to do what's right with thething.
I can't look up, Will, in my my platform.
You're this you know, you're disassociated fromyour data.
I can ask you some questions at the beginningof the study, but outside of that, you own the
data.
You can revoke access to everything, and itwipes your data from the system.
So from user privacy standpoint, that's onething.
(17:05):
But from the business side, I think it makes alot of sense.
We have to better understand the people thatwe're designing for so we can design the right
thing.
That makes a lot of sense.
Kudos to you for thinking of the privacy side.
So for all our privacy experts out there.
And we, as Spot Migration, definitely approveof at least the concept around how you're
(17:29):
looking at data, very of you.
Well, I that GDPR ish, you you mentioned that.
I actually work with Amsterdam.
I hope I said the g right to please my Dutchfriends.
With the GDPR expert over there because one ofthe early explorations we did was back in the
city of Amsterdam, which is which is just afun, like, reconnection.
(17:50):
So, yeah, we we went into it, and and theychallenged us pretty pretty heftily about how
we approach data privacy and and disassociateit from the person that's sourcing.
So, again, when you have a cool thing, peoplepeople like to be a part of it, and they they
were really good partners early as we wereexploring this.
(18:10):
What what have you learned?
So, obviously, we talk about differentstressful internal environments.
But, like, when we think of, you know, if it'sbike lanes or if it's other transportation, you
know, ways.
What have you seen?
Obviously, you felt it right when you wereoverseas.
You you just felt like, is this so pleasantcompared to in my my home base?
(18:34):
You know, where it's like, I know the roadsbetter.
I know everything better because it's my theroute that I've, you know, made to take and all
these different things.
What is what are those big factors that do makethat difference?
That like, oh, wow.
This is these are huge aspects that we can justacross design make changes in.
There there are some commonalities like,separation between uses, green space,
(18:57):
typically, people people like.
There are trade offs that's costly, you know,to do some of those things.
But, it's weird because it's very it's verylocale based.
Like, if you're in Arizona
Mhmm.
You we can't put the same things in Arizonathat we do in in Pennsylvania.
Right?
They they won't grow.
And so it's very locale based.
There's there's different tolerances of stressbased on your locale as well as the background,
(19:20):
what you've been exposed to up to the pointthat you're collecting data.
So that part has been eye opening for me.
But, yeah, there's some commonalities that thatthat track, but then you can optimize it based
on the locale and the expectations of thosepeople in a specific area.
That's cool.
And one other thing I thought that was reallyinteresting that I learned from it, let's say
(19:40):
so I I had this happen to me.
I got hit, by a car.
Every time I passed by that, I didn't have anyproblems for for years.
Got hit, and and the woman could not have beenkinder.
It was it was a silly mistake.
Every time I passed by that spot for up alittle bit over a year, there was remnant
stress.
So that started opening my eyes to meaning,every time I passed by there, even when I by
(20:05):
myself, there was no one around me.
I still remember that experience, and it stillcaused me stress.
And the the the thing about that is when you'reoperating in an environment that you are
stressed out, you are more likely to makemistakes.
And so that remnant effect of stress waspowerful because what happens if I don't do a
good job or, you know, as an architect, wedesign a place that that does not resonate with
(20:29):
someone in a positive way.
They might start avoiding there's avoidancebehavior that that that comes with that initial
experience.
And so those those are things that now we'rewe're starting to talk about and think about as
designers.
You know, we we, again, going back to Will'sinitial question, what's the value?
Well, if we know that people are not gonnarespond well to something and we can verify
(20:52):
that in a simulated environment using, youknow, goggles and, and some some test data from
stress, Let's do that before we build anything.
So we can go we can again, there's so manydifferent directions we can go with it, but
it's just there's lots of little neat tidbitsthat have been tied to just that general
exploration of of people and spaces and places.
(21:16):
Because then you could also take it a stepfurther.
You mentioned the nurses.
Like, you could say, oh, and I'm not suggestingthat we should, you know, be mining everyone's
data if that's inappropriate or not.
But in regards to, hey.
You're okay with this.
This is for your betterment.
Awesome.
You are stressed to this level.
Therefore, we have, like, a a chill out room ora room for you to go kinda reset in.
(21:40):
And, we want you to do that or construction.
Right?
So, obviously, a construction site,extraordinarily stressful.
You have hit a level of stress that it actuallymakes sense for you to just step off the job
site for a second, take a minute Sorry.
Get get your composure, get back, you know, toeven or level, and then, you know, go back to
work.
And we're if using something like this allows,you know, alerting and things like that to
(22:02):
happen, then that's amazing.
Yeah.
So we we're still doing new explorations.
Actually, we call our nurse stations the chillout rooms trademarked
by us.
Nailed it.
But what were things like that again?
Business, you know, business dollars.
If you can save a nurse from stressing out andleaving, well, costs about $250,000 if I'm an
(22:25):
operator for of a hospital to replace thatperson and get them geared up to a point where
they're they're, you know, adding value as partof the staff, as part of the staff.
So there's all kinds of other things that,unrelated to design, but that that clients may
care about that I think some of this datastarts to to shed light on.
It's cool.
(22:46):
Thanks.
That's a what We got so many other cool stuff.
I was saying, so what else are you innovating?
You you know, that's if that's not enough.
So one cool thing I'll tease, and we're we'regonna announce this, via LinkedIn.
You'll see it, here in a little bit.
But, part of the part of the process gettingdesign off, you know, off the ground when our
(23:06):
clients have an idea, they talk to their teamsabout something that they may wanna design, and
then by the time that we wrap our heads aroundthe scope and go back and forth with them, it
can be a it can be a pretty not cumbersome, butit can be a challenge sometimes.
Right?
Because they have a preconceived notion, theyhave aspirations for a place that they wanna
create, and we have an interpretation based ona phone call or an email.
(23:29):
So we're breaking down the barriers and andcreating we've got a a gaming engine platform
that we've developed called Daedle.
It's my team loves Greek mythology, and so allof our when they're in development, we've
always picked Greek mythology kind of thingsfor, you know, to call them.
Daedal is after Daedal is the architect of thelabyrinth where there's all these weird pieces
(23:51):
that can fit together in lots of differentways.
So we're we're gamifying the early designphases of of how we scope projects and letting
our letting our clients come in and andcocreate with us.
So, for instance, let's say you're designing anew, a new industrial facility.
Right?
You're you're making batteries.
(24:12):
So you have a bunch of raw material hitting asite.
How much?
What's the volume of that?
Okay.
Well, the next step is the cathode and anodeplants.
You know, they they split up the raw materials,and they do two different things.
Well, how how much of that can these places howmuch raw material can these all go through a
process and turn it into something else?
What's what's the the cadence of that?
(24:32):
What's the volume?
What's the demands on the energy grid?
All of those things, we start to tack on layersof information.
And it's doing a simulation.
It's showing a simulation of those things.
And we have historic data about a lot ofprojects.
Right?
Because we've been in business, you know, forfor a long time.
We start to layer on.
Well, the last time we did an anode and cathodeplant, it it took this much area.
(24:53):
So now we're starting to rough out the spacethat supports a function that's defined by a
process that our clients are putting in withthe game engine.
And so, again, if you think if you're if you'redorky too, you've seen things like Factorio or
Satisfactory.
We're we're building things like that thatallow us to extract better information about
the processes, that allows us to better designthe places that support those processes.
(25:19):
So that's that's another thing, and and we'rewe're going one farther.
We've got, we're calling it a fidgetalexperience.
So you can play the game.
You can do it online, you know, or whatever,but you can also be moving blocks around that
are designated with these RU codes.
They're like a simplified QR codes withcomputer vision, with projection mapping where
you're moving a loading dock around, and itshows you on a real time adjacency matrix.
(25:44):
Hey.
Your loading dock is too close to the office.
Here's some negatives that you get from airquality as a resultant of that move.
So why don't you move back away?
And so the projection mapping is is making itred.
Like, there's a violation here that you maywanna consider.
So that's that's a cool thing.
I'm I'm really excited.
Super cool.
Oh my gosh.
We hadn't talked about that, but that's anotherthing, my team's been been really working on
(26:09):
hard, and it's it's looking sharp.
It's looking really good.
So that's another thing.
That that, again, that's super secret.
You guys get a teaser.
That's the first time I think I've talked aboutit externally, so you guys you guys get it
first.
But that's Oh,
there's here first.
Well but so let's do that.
Let's do that first with our clients.
You want me to tease more stuff just for fun?
(26:30):
Sure.
Yeah.
I think our listeners will gobble it up.
Well, it it stopped me because I I mean, I I'mseriously I can go because I love this stuff.
So that's step one, like understanding theprocesses and understanding the space that our
clients expect us.
Well, what's next?
We have a massing.
We have a rough area, you know, of spaces nowthat we have to accommodate, and we might even
(26:51):
have some rough layout of this the space.
Well, we're doing a real time rendering engineon top of it.
So as the the digital cameras are picking upthe layouts or we're having the the digital
pass off from our game engine, we have, what wecall Gia Imagine.
Gia is our internal chatbot.
She she is it's it's we've engendered her.
(27:12):
She's female because she solves all ourproblems.
But we have modules for Gia, and it stands forgenuine ingenuity assistant.
Genuine ingenuity is the Gresham Smith tagline.
Harkens back to the early you know, what I wastalking about.
It's our a little little aside.
Genuine means we care about we care about ourwork.
We care about the impact, so we do it in moreingenious ways.
(27:34):
We we tackle things in in kind of moreintelligent, more smart ways.
Gia imagine starts to pick up thisinterpretation of amassing and do real time
verbal input.
So we can say, hey.
Here's a reference image.
I really like the glass on this panel, but Iwant it to be darker blue, whatever.
Gia imagine is taking that and doing a realtime rendering at the same time.
(27:56):
And that's the mid that's the middle piece.
That's working too.
So the last piece of that is another iterationof Empath.
It's called Empath React.
Just like I'm I'm seeing your faces.
Again, keeping the design process very human.
We're good.
You know, I'm seeing you both smile.
That's great.
That makes me feel good, but I'm doing thatbecause of a combination.
I'm tracking I know your mouths are a littlebit agape.
(28:18):
I see your teeth.
I see you have a a slight smile.
I know you're squinting, so you're interested.
Those are all things that we do every day.
Empath is is something else that that I'vebuilt with with my team over the past cup maybe
three or four months.
It's been fast, but it's an emotional model.
So it looks for known points on your face.
It looks at the math between those known pointpoints and the angles that they change, and and
(28:43):
it picks up, are you responding well to whatI'm saying or showing you?
And so that's the the last layer of this is andfor it does verbal sentiment analysis through
in real time too.
So as I put down a selection of lightinglighting or furniture in front of you, You
could say, oh, I love the brown couch.
I hate that light.
(29:03):
That would go good with this.
It's it's listening to you, and we're passingthat back to g imagine.
So it's swapping things out based on yourreaction.
Oh.
And The design elements.
Yeah.
So that Are
you talking about, like, even, like, microfacial expressions?
Yes.
It's picking that up?
Exactly.
That's exactly Oh, jeez.
Okay.
That is wow.
(29:25):
That that's impressive.
Well, the now the tie back is you you justheard, like, the beginning part of we wanna
design better places for people.
We wanna make it fun for our clients, and wewanna get better information out.
We wanna show them how we augment, how we howwe're not just replacing our our professionals
because these these are the the people that Iwork with.
(29:47):
You know, we have 1,200 employees at GreshamSmith.
They're brilliant.
I don't wanna replace, you know, their talent.
So let's augment.
So we're we're making the the creative partfaster, and then we're using again this this
next iteration of empath, empath react to pickup, you know, slight variations and changes in
facial cues and verbal sentiment analysis todesign the right stuff faster.
(30:10):
And so we're we're lumping all this togetherinto the Gresham Smith design the Gresham Smith
design lab.
So So that that's close.
If you so if the term hasn't been coined yet,you heard it here first.
It's not vibe coding.
It's vibe architecting.
Oh.
I'm gonna submit for a trademark, and I'll I'llgive you credit, Will.
(30:32):
Appreciate that.
That's just a glimpse, and I didn't wanna spendtoo much time on just the impact because it
sparked that little bit of a challenge, likehow we're approaching design.
We could be doing that daily.
We could be evolving and making this process somuch more fun and so much more enjoyable for
for 1,200, you know, Gresham Smith architectsand engineers and program managers.
(30:57):
Let's challenge all of that.
And and, you know, Will, I know we talkedbefore during introductions before about
digital transformation.
I hope now you see this is what I'm talkingabout.
We do all of these things manually.
Now we have tools that we've either developedor that we, you know, can guide to to really
evolve what Gresham Smith is and and define apath for us that that is the future of design.
(31:21):
And I think we're we're on to something.
Is automation part of your scope as far asinnovation goes?
It is.
Yeah.
It absolutely is.
So then let's talk a little bit for a second.
Mhmm.
What's the secret to automating things?
(31:41):
So it's it's really it starts with the team.
I don't know the so let's let's take forinstance, a mechanical layout of a plant.
Well, our practitioners know how that goes.
They can do the manual process all day long andin their sleep almost.
Me as my background, you know, as as Justinsaid, his background is transportation
(32:03):
engineering with this weird overlay oftechnology.
I will never learn more about that process thanwhat they already do.
So it really starts with the team.
They have to identify things that they wouldlike to target for automation, and then we are
very reliant on them to spend the time tomanually map all the processes associated with
that manual working through it.
(32:24):
Woah.
Woah.
Woah.
Hold on.
Hold on.
You said you rely on them spending time.
That's right.
By any chance, is he, like, your middle namechief cat herder also?
It it can't can't feel like that, but you wouldyou would be surprised, when they get it,
they're really willing participants.
So it's less herding cats and moreorchestrating slight chaos at the beginning.
(32:50):
K.
That actually talk or speaks to the culture.
Right?
Because, like, it is the idea that we are goingto, create better things.
Right?
Like that, you know, and however you chalkchalk that up.
And that's not just the innovation team.
That is across the board.
We are gonna create a better client experiencedoing that.
So if we automate things that will create thatbingo bango, here it is, which, is a beautiful
(33:14):
thing to see when it comes to a company of yoursize to have culture embedded that deep.
I I love that you brought this up.
Our our CEO, Rodney Chester, has been reallyinstrumental in in helping to guide with all
the things that we're talking about today.
He set this up for success, and you can see itwhen you walk through the on people's faces.
They're excited.
And Rodney, again, I I don't wanna I don't wantthem to think I'm bragging on them too much on
(33:37):
the on the podcast, but I'm pretty sure.
But but he's come out and said, I wanna makesure that we have the best culture of any AE
firm out there.
And then he turned to me and said, we have tomake sure that when people choose to spend
their time at Gresham Smith, that they have afull long lasting career that they feel fully,
like, invested and they're getting somethingout of it too.
(34:00):
Well, to me, I can't turn on anything that'sgonna short circuit that.
And so going back to automation, the reason Istart with the team is I really want them to be
guiding and leading and picking the rightthings to automate.
There we could probably automate the world.
Like, there's so many things we could beautomating, but it short circuits.
It either creates knowledge gaps.
It creates, you know, issues, for forbillability on things.
(34:26):
There's all kinds of other things that we haveto be cognizant of, and it there's trade offs
in all of this.
And the other thing too, we're we're, you know,we're stamping things.
Architects and engineers certify our work, andand I'm very cautious of automating something
that could be interpreted as, skipping thestandard of care.
That's that's a nonnegotiable for us.
(34:46):
So it's all of those things yield somethingvery positive.
But, yeah, it starts with the culture.
I love that you brought that up, Justin, andthat's something that's really near and dear to
to our our executive team at Gresham Smith.
So to go, you know, a step a step further intothe automation side
Yeah.
(35:06):
How is it affecting work?
And you said there there's some you described aphrase, and I want you to kind of maybe either
zoom out or zoom in.
I'm not sure I'm not sure which direction youwanna take this.
But when you talked about sort of the care andmaking sure that there's certain pieces that
(35:31):
are not skipped, are you saying that there arecertain pieces that you could automate, but
you're, I guess, making a conscious decision ofnot automating?
Absolutely.
Yeah.
Yeah.
Talk about that a little bit.
Yeah.
So, especially when you have, like, onboarding,you bring a new engineer or architect on.
You want them to learn the actual steps.
(35:52):
You want them to learn the process for how tothe Gresham Smith way that we approach design.
If we skipped those things, if we automatedthat early experience, you're in effect
undercutting their ability to grow as aprofessional.
You have to be we have to be very careful aboutthat.
So, yeah, we're not we could be automating alot more that we're we're purposely not because
(36:13):
we wanna build great designers, architects andengineers, and and everything else.
Consultancy services we do.
We don't wanna short circuit their ability tofeel fulfilled, to grow as an individual and
professional, and and understand the decisionsthat they're making on a project have real
implications.
So, yes, we're not we're not automatingeverything, but we do choose to automate some
(36:35):
things that are highly tedious that drivepeople insane.
Absolutely.
But we still walk them through what it's doingtoo just so they're they're aware.
Are you able to share, any of these, we'll calltedious things?
So just to give an idea to people that are outthere, like, what you think is worth automating
(36:56):
and spending the time?
Because it takes time to automate.
Right?
Like, automation is great, but there's a lot oftime that's spent on doing the automating
before you actually get reap the rewards.
Yeah.
So so there's several.
One thing in particular that I've been reallyproud of is there we have some really senior
level folks that have built these massprocesses in Excel to do calculations and
(37:19):
checks on things like electrical layout andsupportive systems.
And the neat thing about this is it's it takesa lot of time.
You get a you get a a plant layout.
Again, we'll go back to industrial sites thatare very complex.
They have lots of different manufacturers thatthat you have to support things moving in that
that do things with raw material or or part ofa process.
(37:40):
Just the electrical side of that.
Let's say one of the equipment manufacturersswaps out something.
That changes everything.
It changes the the entire electric grid and allof the panels and all the support mechanisms
behind that.
Every time one of those change, and and thatchanges a lot.
So you take a junior engineer that is manuallygoing through the same process thousands of
(38:04):
times, and you automate it and say, hey.
We just saw you swapped out this.
We can read because it's a smart object.
What the load is, we're gonna go all the wayupstream and change all of the panel
configurations, And here's the output, andhere's a report of what was changed and why it
was changed.
Check check this.
Do you agree with it?
Yes?
Okay.
Well, you just saved literally two thousandhours.
(38:28):
Oh, jeez.
And I'm not
I'm not I'm not exaggerating at all.
That that's that's exactly how that processwent.
And I will say kudos to to that team.
They invested about a month's worth of time onthe front end to to convert that Excel into a
logic map for us to follow.
(38:48):
And then my team was able to automate it inabout two days.
Okay.
Well, that's worth it.
I mean, so took a month.
So two thousand hours, that's basically ayear's worth of man hours.
A two thousand hours on a on one project whenone thing swap
out.
One thing changes.
(39:09):
Oh, jeez.
Can be the scale of it is in enormous,enormous, and it's tedious work.
Why would we you know, we want people to beexposed to the work, but we don't wanna bore
them to death.
You know, we don't want to get them superexcited to come to work for an architecture
engineering firm and then spend all dayreplacing, a decimal place in BIM.
(39:32):
So but the other thing too, it's not just timesaved.
It's it's better for risk management.
We've we've eliminated manual everything wherewe can on things like that.
So the me getting tired after six hours ofdoing what I just described, maybe I do miss a
miss a number.
Maybe I put a nine instead of a six.
(39:52):
Well, because it's automated and because weverified the automation process, we we get
better delivery that's less risky.
Yeah.
Removing the human error, essentially.
That's it.
Yeah.
No, that's and morale.
Like, yes, x amount of time.
Sure.
But like when you're doing something that suckslike that and you're doing it for that long,
(40:14):
it's like, oh yeah, that's such adeterioration.
And then you're unhappy.
And then if you're unhappy, those around youfeel bad, and it goes you know, it chunks on
and on and on.
I don't know how many modern interns would bewilling to move a decimal point, and spend two
thousand hours, doing that, across an entirebuild.
Yeah.
You're you're absolutely right, Will.
And and when you think about, okay, an internthat has two choices, they've got they can come
(40:37):
to Aggression Smith and have all this coolstuff and support and be a part of, you know,
evolving, and and we we let them.
We have junior level folks on our futures labis what we call it.
They give us ideas and we we evaluate those.
They can be do that or they can go and do allof these manual processes at another firm.
Well, guess who's gonna win?
And there's a war for talent like that's a realthing.
(40:58):
Finding enough folks that really wanna do whatI'm describing to you and have have a positive
impact on design, it's still tough to get them.
So yeah.
What has the, essentially the investment intechnology automation, whether it new new
programs, new innovations, or just thesenondesimal moving automations.
(41:21):
Like, what has been that impact when it comesto hiring and retention?
Well, we're we're growing.
I mean, we're that's the thing.
We have, I think, a 90 something.
I can't remember and I don't wanna quote, butit's north of 90.
It's way up there for employee satisfaction.
Our retention rates are very high.
We have very low turnover.
And typically, when we get to offer stages, wehave a lot of success, not just for innovation,
(41:45):
but just in general, the overall culture, howwe approach approach design, and the accolades
we get for doing good design.
So it's it's a innovation's a piece.
It's not the answer.
It's a piece of the overall value that that afirm like Rushing Smith can bring.
But, yeah, it it works.
You're you're so humble, Mike.
You should be like, it's all innovation.
(42:06):
That's the reason we we exist.
It's not.
And I think if if I was on here saying that,you should not be talking to because
I can care.
I'm joking.
Yeah.
No.
No.
No.
I get it.
But, yeah, it is a it is a piece of it.
But there's some tried and true things that arealways gonna exist in design that we don't
wanna disrupt.
But there are other things right fordisruption.
Well,
(42:29):
we don't wanna, you know, we don't wanna diginto any future innovations and, you know,
however, are all innovations is it seeneventually at the client end?
Would you would you say that the client sees itor customer, however you say that?
(42:50):
Historically, we've done a a bad job of that.
I I'd like to sit here and say everything'sbeen smooth sailing.
It's not.
We're we're rebuilding this plane as we fly it,as some would say.
We're figuring a lot out as we go.
And I would say, historically, we've done a abad job of that.
Now, though, we're taking a different tackwhere we are since we've kind of established
(43:12):
the space of trust with our internal marketsand corporate services inside of innovation,
our markets are starting to say, hey.
We're gonna have clients come and evaluatethis.
Is that cool?
Yeah.
That's great.
And our clients may say, man, I wish you woulddo this and this, but we can do that.
But having them come to the table and and Ithink us being willing to let our guard down a
(43:33):
little bit and and I like the
word
cocreation.
My my team will make fun of me because Iprobably overuse it to them, but I think it's
important.
And we can do that not just with design.
We can do that with these platforms.
And so it's been fun to to tease these out.
Like, all the things that I've just describedto you, clients have seen, And and they're
helping to shape, and and they're gonna be partof the early adopters.
(43:56):
But but we're building it kind of together.
And so, yeah, we've gotten a lot better at atbeing willing to to pull the veil back a little
bit, let let our clients see rough ideas beforethey become, like, tangible.
So, yeah, I I think that's gonna continue.
And and I think there's gonna be a lot more ofthat that model moving forward.
(44:17):
You spoke about Gia as sort of your internal AIchat.
Mhmm.
I'm assuming you were part of that process.
Yeah.
For companies that haven't actually dug in.
Mhmm.
Or they're at the very beginning stage of theirprocess around choosing the direction and
(44:44):
policies and all that.
Mhmm.
What are some learning lessons that you guyshave learned Mhmm.
And maybe if you'd had to do it all over againor would have just either been better if you
sped things up.
Right?
What are some of those learning lessons thatyou learned as you were implementing
Sure.
Well AI.
(45:04):
Yeah.
It's a great question.
And we were we were really early adopters.
Again, I'm fortunate to have, you know, anexecutive team and a board that that supports
explore or exploration early.
The things that weren't well was we went out ofthe the gate with, we had a governance policy
that we we drafted almost immediately.
We had a usage policy as we start to adopt newtools, even before we adopted them.
(45:27):
As we start to adopt new tools, here's what weexpect you as a user to kind of concentrate on,
which is not skipping the standard of care, notobviously using the tech for bad stuff, not not
skipping parts of the process.
Then we started digging into specific tools wewanted to use.
The first version that we unveiled of Jia wasresided in Teams.
(45:48):
This was two yeah.
Two two years, almost three years ago, so itwas early.
We put an early AI agent in Teams.
We coded it ourselves, and it was okay.
You know, it kind of unveiled that, hey.
We're we've adopted something early, but it hada lot of hiccups.
Like, the history was funky.
(46:10):
You couldn't reference things without manuallygoing back and doing kind of the the now the
current version of of most platforms.
You go back and forth with chat.
The so that was a lesson learned, but that wasthe option.
The next lesson that I would say is really diginto the buy versus build.
So for Gia, which is an internal platform thatwe've coded, it actually has more functionality
(46:31):
than OpenAI's offering.
We built it for about $12,000 total, and ourmonthly reoccurring cost is between 5 and $600
on API usage.
APIs, we did that specifically because the datathe data privacy inside of an API connection is
is secure.
It's protected.
(46:51):
So if we turn the same thing on using aprebuilt system like an OpenAI or Microsoft
Copilot, which, you know, there there arepluses and minuses to that, that would have
costed anywhere between 250 and $450,000.
But we have expertise.
We have the expertise internally, and so webuilt something that's a really great use of
(47:13):
dollars.
That first chatbot, the first the v one, whatwas that to the did the whole company get
access to that, or was that just innovation?
Okay.
So whole company did.
What then was the use case for?
Like, what what were or what were people usingit for?
Or what people what did people think they coulduse it for?
(47:33):
Maybe some Yeah.
Question.
Pretty much anything.
I mean, it was all over the place, and we werepretty open and honest.
That's another thing, advice I would give.
Lesson learned is be open and honest that, hey.
We're watching.
Don't put Yeah.
Terrible stuff in this.
You know?
And and we trust our folks.
That that didn't happen.
But we told people, hey.
We're watching this, and we did it with youknow, what we wanted to do was if we find a
(47:58):
bunch of people using it for design processesor challenges.
Mhmm.
Let's invest more of our time from theinnovation side in IT and MIS on focusing on
those types of of agentic systems or the customknowledge bases.
If people were using it for recipes, well,we're probably not gonna invest a lot of time,
and there are people that use it for recipes,obviously.
But but we tracked it to make sure that we'reinvesting our time in the right types of
(48:22):
knowledge bases or or add ons for our internaloffering.
And we launched Gia chat version two, and we'vegot generative image and video and everything
now too.
All all Okay.
In in house.
That helped to guide, you know, the types ofthings and system we wanted to build.
And so we do have all of those custom knowledgebases.
(48:42):
We've got, automation scripts that go back andforth.
We've got NCP connections now inside of of Giachat that connect to other things.
I can't talk too much about it.
There's a lot of there's a lot of secret stuffaround that, but it's doing a lot more for us
because we were open and honest with the peoplethat we're engaging with the first version.
So that that, I think, is another lesson there.
(49:04):
And the last one, lesson learned going back tothat question, I know you've asked other
things, but this is an important one.
Just because you turn on something or youinvest in something and you expect everyone
else to be just as excited as as, you know,maybe maybe us on the phone.
That just doesn't happen.
And so we we started coming out of the gatewith, here's use cases, and we let those people
(49:25):
that develop the use cases own them andrepresent them to the firm.
We have we have, innovation host quarterly,like, firm wide calls where we showcase a bunch
of cool stuff happening.
Again, trying to get more people excited, butthe use cases and then the training videos is
you gotta do that because not everyone is goingto be as curious or explore the same way that
(49:48):
that I do.
So, yeah, that's that's
Does g o two now you just prompt g o two tomake those videos for you?
Because it's just like that.
So we we have started building commercials.
You'll see all the things I teased earlier,like the the big new we've got commercials that
we've developed.
Yeah.
They've got voiceovers and look like we'vehired a production company.
(50:08):
Yes.
Nice.
Yeah.
Also, looking for recipes could be a goodthing.
There could be a company cookbook.
That that's like a document that could live onyour SharePoint.
I I love I love,
you know, cooking and and things like that.
And and ultimately, whatever.
You know?
I'd rather somebody be curious about it andtinker with it than oblivious to it.
So
(50:29):
I also think that in, how how it's set up, butlike getting them to just use it, right?
That's the biggest hurdle because then onceonce, and this is what certainly we've seen
when somebody's interested about AI, it's like,well, are you, what are you doing with this?
So I've not done anything.
It's like, okay, Well, just start playing withthe chatbot.
(50:49):
Right?
Like, just start start there.
And then what will happen?
You'll start getting a lot more ideas of like,oh, this could work for the thing I'm dealing
with in business.
Like, oh, but like this totally I could usethis this way.
And it's like, yes.
Exactly.
And that's how innovation works.
Right?
Like, that's the just explore the next thing.
It's it's fun.
(51:09):
Like, last week, we gave a presentation,current use cases to our owner's group, and you
saw light bulbs just going off across the room.
They're like, oh my gosh.
I should be doing this.
It would save me so much time or I could beputting together such a better proposal
response.
And so that just level of exposure to somebodydoing something really unique and saving a
(51:30):
bunch of time or producing something betterwith the help of this augmenting tools.
It's just making us better.
Yeah.
Light bulbs across the room.
It was great.
That's awesome.
Yeah.
Justin, I think that we are at a good point togo to our last question because otherwise, we
(51:52):
are gonna go on like Joe Rogan for four morehours.
I don't know.
I have it in me.
Yeah.
I don't yeah.
You're, East Coast time.
Yeah.
Awesome.
So, Mike, we love to ask this question toeverybody.
If you could go back twenty years, what advicewould you give yourself?
(52:12):
Oh, that's a really good question.
What advice would I give myself twenty yearsago?
I would say, as with anything, I struggled, youknow, when when you're intelligent, when you
when you can process lots of things reallyfast, you have lots of big swings.
You got lots of dips along with the peaks.
(52:33):
I've talked mostly about peaks today, and Iwould say it's worth it.
Like, stay curious.
Keep digging.
All of those little moments that I spentteaching myself Pascal and something that I
didn't think would amount to anything, but Iwas just curious about, they all add up into
something really cool as long as you pursue it.
(52:54):
And I think that marrying up something that Ilove to do, which was I wanted to influence
places, and I've married it up with that lovefor technology.
Pursue that passion.
That that that the only reason I'm able to talkto folks like you and actually feel good and
confident about the the future of my of thiscompany, Gresham Smith, is because of that.
It's because I've I have been very fortunatethat I've been in a position where I can marry
(53:18):
up multiple different passions and and helphelp help a lot of folks.
And I would say, I didn't mention this before,but it works.
We just won, Fast Company's one of the mostinnovative companies in the world.
That's prestigious.
Well, it's but it works.
And so it's, again, it's just that that staycurious, marrying up the passions, and pursuing
(53:41):
it.
It it it ends well.
So, yeah, I'd say that.
Last words.
That's that's This
has been a ton of fun.
If people wanted to get ahold of you, what'sthe best way for them to do that, Mike?
I would I would direct them toaggressionsmith.com or or or find me on
LinkedIn.
I'm pretty active participant on on that.
So, yeah, happy to engage with folks.
(54:02):
I'd love to hear their thoughts about some ofthe new innovations coming from aggression
smith.
Awesome.
Awesome.
Awesome.
I'll make sure I put, your LinkedIn, and thewebsite in the show notes.
And is there anything else you'd like to tellthe people before we say goodbye?
I would say you're getting ready to see so muchchange.
It's gonna be a really exciting time for notjust innovators, but just I I I listen to your
(54:25):
podcast.
I love what you all do, and there's a lot offolks that really care about about all of these
things.
And I think you all do a great job, so there'sgonna be so much more opportunity to capture
this and to do cooler and more impactfulthings.
Just get ready.
It's just it's gonna be a lot of fun.
Awesome.
That's awesome.
Thank you.
Thank you for listening, Mike.
(54:45):
We appreciate it.
I appreciate the opportunity.
Thank you guys so much for hosting.
Awesome.
Well, listeners, I hope you had as good of atime, and maybe mind was blown with, the
innovations that are coming out of GreshamSmith, as me and Will are.
But until next time, adios.
Adios.
You guys.
Thanks for listening to building scale.
To help us reach even more people, please sharethis episode with a friend, colleague, or on
(55:11):
social media.
Remember, the three pillars of scaling abusiness are people, process, and technology.
And our mission is to help
the AEC industry protect itself by makingtechnology easy.
So if you think your company's technologypillar could use some improvement, book a call
with us to see how we can help maximize your ITcybersecurity strategy.
(55:35):
Just go to buildingscale.net/help.
And until next time.
Keep building scale.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!