Latest / Wired to Build / Hands-On to Hyperscale | Designing Smarter Data Centers with Kaylen Bushell

Transcript

Hands-On to Hyperscale | Designing Smarter Data Centers with Kaylen Bushell · Wired to Build

0:00
Hey, what's going on everyone? Welcome to Wire to Build, a
0:03
data-driven technology focus podcast digging into how
0:05
building owners and operators use technology to manage the
0:08
full life cycle of their projects.
0:10
I'm Nick Carabella. You're a showrunner and curator,
0:12
and I'm excited that you're joining us today to talk about
0:14
the people that drive the change in our industry.
0:17
You we're coming back for a second episode with Kaylin
0:20
Michelle very special guest here from the Avocado team and we're
0:24
going to be digging into data centers and her journey from the
0:30
Coast Guard all the way through to breaking into the industry
0:35
and titled onto hyperscale. So we hope it's going to be a
0:38
fundamental with you. I'd like to welcome Christy back
0:41
to the show. How you doing today?
0:42
Hey. Nick doing great.
0:44
How about you? Doing well.
0:46
I'm, I'm really excited that Caitlin's here.
0:49
I mean, and I think now she might even be top 9.
0:51
I mean, you know, we keep doing this a little longer.
0:54
She made crack top eight. I don't know.
0:56
But Caitlin, what would you remind folks for those who maybe
0:59
didn't listen or have short memories like me, remind folks
1:03
about who you are and and what you do.
1:06
Yeah, yeah. I'm hoping I crack into the top
1:08
five by the end of this episode. So I'm going to really lay on
1:11
thick. Yeah.
1:13
Kaylyn Bushel. Yep.
1:15
Coming to you from Avocado. Been here for three years and
1:19
have spent the last 12 years in the data center industry.
1:24
Is kind of in various roles, but mainly in the operational and
1:28
construction space and kind of serving as like a a translator
1:33
of sorts. Super tech for Part 2 by the way
1:35
guys. So, so are we.
1:38
I appreciate that translator role.
1:40
Not everyone can do it. And it's something that is
1:44
actually a lot of value that it's hard to convey.
1:46
And it's not like you go to school to become a tech to non
1:50
tech translator. So I want to dig in on that a
1:53
little bit because I think that's such an important part of
1:56
the success that you're having that we're having and, and in
1:59
the industry of getting things done more, more effectively,
2:03
more efficiently. So how are you helping connect
2:08
dots for folks across those various systems that you're
2:12
working on? So running fiber and tracing
2:15
circuits and kind of doing all these things.
2:17
I'm already over my skis and talking about that.
2:19
That's not my world. So how do you how, how, how does
2:23
that go for you and how do you help manage that especially with
2:26
growing complexity of data centers?
2:29
Yeah, yeah, I'm like AI am a lover of metaphor.
2:35
And so I bring that with me. It's like my favorite tool in
2:39
the tool belt is kind of making real world connections.
2:42
So you know, as as you know, in like, I mean, it's it's true in
2:45
in construction as well. And honestly, I'm sure it's it's
2:49
true in in the marketing world where you know, you when when I
2:54
need to convey something from an upstream.
2:57
So when you think about the life cycle of data centers, you know,
2:59
we have this, we have this period of like site selection,
3:03
due diligence, and then there's like this period of like
3:08
permitting and pre construction. And then there's like a design
3:11
phase that sort of overlaps all of that.
3:14
And then then we go into, you know, construction and then we
3:17
we turn over. That's a very, very painfully
3:20
oversimplified explanation, but just just to paint the picture.
3:23
And one of the things that I've prided myself on is being able
3:26
to a complaint. For example, let's just I've
3:29
been in a design meeting where they wanted to implement this
3:33
new design, put the a relief fan in an overhead plenum.
3:37
So it's like in a, in a hole in the in, in like avoided space
3:42
overhead. And the, the goal with the data
3:44
centers is always to get a smaller, smaller footprint,
3:46
right? We want more power, more
3:49
cooling, but like with less physical land and coming from
3:53
the operational side, you know, I'm like, my mind is going, how
3:57
do we change out the belt on that thing?
3:59
Like it's inside this, this overhead.
4:01
And there's not like a saffolding.
4:03
There's not like somewhere I could physically walk inside of
4:06
there. Like, so am I supposed to?
4:07
And then I like, I supposed to bring a lift in here, but
4:10
there's like stairs that go up into this plenum.
4:12
Like how am I supposed to get that up there?
4:13
I have to bring a ladder. You know, these are just going
4:15
through my mind. I'm like thinking through all of
4:17
these and, and you think to yourself, surely someone has
4:20
thought of this. I cannot be the smartest person
4:23
in this room that is bringing this to the forefront of
4:25
everyone's mind. And and I asked the question,
4:28
you know, how do we how do we replace the fan or how do we
4:30
replace the belts on these or replace the bearings or grease
4:33
the bearings on these on these relief fans?
4:36
That the guy that, you know, he was a design engineer, like
4:39
probably a mechanical engineer. And he's like, honestly, I
4:41
didn't think about that. And that was kind of a big
4:44
moment for me where I said, OK, like it's not that they didn't
4:48
care, but but that's not their purview.
4:51
That is not their goal is to make sure does this fan remove
4:54
enough physical air to remove the heat from this building?
4:58
And it surely probably did that. But you know, and so when I
5:02
think of like myself as a translator, what I try to do is
5:05
like stop assuming that someone is thinking about, you know,
5:11
your area of expertise or your Ave. of, of knowledge and, and
5:16
try your best to do a little, a little bridging, if you will.
5:22
Yeah, I well, and I think that that's that's so important is
5:25
and something that's easily to overlook it because as humans we
5:29
have our unique perspective and we tend to think like everybody
5:32
is probably approaching this from our own world view, which
5:37
is like the you could not be more wrong and like be having
5:41
empathy for the designer who and many times early in my career, I
5:44
would make the same mistake. My job was to get it in there,
5:46
make have it work. And like I never really thought
5:50
about the maintenance side of it.
5:52
And it's so common early in an architect's career to be like,
5:56
what does it mean to change this light bulb or what does it mean
5:59
to maintain that belt? Can someone get in there?
6:02
Do I not need to rent a scissor lift?
6:04
Does this become like the most expensive thing for me to do to
6:07
change one light bulb on a building?
6:10
And then through I have to do that 100 more times and now I've
6:13
created a full time job for light bulb changer, which
6:16
happens. And it's so interesting 'cause
6:21
that's where, you know, I think the difference is of generalist
6:26
and specialist and the importance of both roles get
6:31
traction in the construction in the street.
6:33
And so I know we talked about like you starting as, you know,
6:37
a welder and kind of growing through those ranks.
6:39
How you know, in that with the 1 little anecdote you shared, you
6:43
know, what is that relationship that that you try to establish
6:46
as journalists to the specialist you work with and helping
6:50
create? I don't know what's called like
6:52
the the whole solution if you. Yeah.
6:55
So what you're asking is, you know, taking like in and maybe
7:00
in an example where, you know, we get a little myopic like in
7:03
a, you know, like an electrical problem.
7:06
Let's just maybe we start there. But then, you know, maybe that
7:10
electrical problem has mechanical implications or it
7:14
has physical implications. Maybe it creates more heat in a
7:17
room. And so is that kind of what
7:19
you're thinking of as like? The cause and effect
7:22
relationship of like trying to solve a problem with laser laser
7:26
surgery as opposed to like figuring out the entire totally.
7:30
How it's all yeah, yeah, yeah, yeah.
7:32
This is really, this is a great question and it's, I think this,
7:37
This is why, you know, data center operators are unique in
7:41
that it almost reminds me of like, you know, the ultimate
7:46
like domino effect or, you know, you, you want to physically,
7:51
physically move a rack. You know, and it's like this is
7:54
something that I just am dealing with with a client right now
7:57
where it's like the, the rack is physically very heavy.
8:01
Like, I mean, these, these can be, you know, some later
8:04
generations can be upwards to 10,000 lbs.
8:07
And you know, what effect does that have on the structural
8:13
integrity of the building? OK, well, now we've eliminated
8:15
raised floors. Now we have to do, you know, a
8:18
slab on grade construction or like, you know, to your point
8:22
with the, you know, we have this one small problem where we need
8:25
to get more, you know, the racks have more power.
8:29
OK, well, do we need to space them out more?
8:31
Does the CFM support this much power in this physical space?
8:35
And so I love this is one of my, the one of my favorite things
8:38
about that, you know, data centers.
8:40
I'm a little ADHDI don't know if you guys have like could have
8:43
diagnosed me on this episode, but I think this actually served
8:46
me really well because what, what I can do is like, I don't
8:50
get too laser focused on anything.
8:52
And So what it what you know, the data center interface
8:55
allowed me to do is like zoom out and think about like the big
9:00
picture, you know, almost like I'll like play a movie in my
9:03
head, like, OK, if we do that, then this is what happens.
9:06
And if that happens, we have to do this.
9:08
And so, and, and I've worked with, you know, a lot, a lot of
9:12
folks in data centers come from the military, a lot of Navy new
9:17
submariners. And I think it's because they,
9:19
they kind of operate in that same world where they have a
9:22
job, but they constantly have to like think about it from, you
9:28
know, the 10,000 foot view of like, what happens if I do my
9:31
job wrong? Like how does that impact
9:33
everyone else? And how's that, you know, bleed
9:36
out? So I don't know if it answered
9:38
your question, but more like validated like your position
9:42
there is like this is a world where we we really can't get
9:46
into silos of like hyperspecificity in like, you
9:50
know, knowledge and operations. Yeah, it just reminded me.
9:53
I watched the Air Force Thunderbirds documentary on
9:56
Netflix. Have you seen that?
9:58
Just I don't. Know if it just came out, but
10:00
they talked, they talked about on their door heading out to the
10:05
flight range. It's called the one core value
10:08
that these have and the Air Force Thunderbirds, they're the
10:10
people who do the trick flights like over stadiums and really
10:16
tight formations. And so they have blind trust
10:20
written on their door. And the entire concept of it is
10:23
that, you know, they're a complete team that has to act as
10:27
one singular unit with such precision that if they're
10:31
worried about someone else's doing their job for even a
10:34
fraction of a second, they could crash their planes because they
10:37
have to be so keyed into what their core responsibility is.
10:42
They get into like how the magnitude of that because they
10:45
fly in that they do this thing called like a diamond formation.
10:48
And if the second pilot doesn't get right into the right spot,
10:52
the air pressures won't be. Where the?
10:55
Tire like they won't get lift and they'll actually crash into
10:58
the ground and die. And it happened I think in 19
11:01
like 83 if the quoting the movie correctly.
11:05
But it's that same thing. I think when you're looking at a
11:08
data center and the fact that they has to have the uptime and
11:11
the the trust that you have to have in the engineers, right
11:15
down to the janitor and the security person.
11:17
And you talked about like how these people are the ones that
11:19
actually end up growing up through the ranks themselves.
11:22
How critically important it is not to understand everybody's
11:27
responsibility and then to trust them completely and making sure
11:32
that they're doing their job so that you can do yours.
11:35
And to me, that's like a secret sauce to high performing teams
11:39
that we see in the data center space of, you know, what do you
11:43
need as the electrical engineer? How does my work as a mechanical
11:46
engineer tie into that so that we can create this whole system
11:51
of trust that helps us, you know, meet what we're trying to
11:55
deliver. And you know, long story again,
11:59
to kind of get where I'm going here is you, you, there's this
12:03
aspect of redundancy, right, that is so critical in data
12:07
center. Something's going to go down.
12:08
You're going to have to even if something doesn't go down for
12:11
bad reasons, you're going to need to do maintenance, right?
12:13
And so there's this aspect of, you know, failure domains and
12:17
tolerances that's so important. Can you tell us or or paint a
12:20
little bit of a picture of what that looks like in the data
12:22
center space and how maintenance teams and building teams have to
12:27
work together to make sure they can meet their up times?
12:31
Yeah, yeah. It's, it's kind of like a,
12:34
almost like a, a puzzle in a way.
12:36
So when we think about redundancy, you know, sometimes
12:41
it's physically and equipment. So there's, you know, again,
12:45
thinking back to, you know, maybe the early 2000 tens when,
12:51
you know, when, when there wasn't sophistication that we
12:53
have today in like traffic failover, which I'll, I'll talk
12:57
about that in a second. But you know, we would some,
13:00
some data centers would have like two utility feeds from 2
13:04
separate transmission lines from 2 separate.
13:07
Like, you know, basically what you're always trying to do is
13:10
eliminate a single point of failure.
13:12
That's like the entire point. So you don't, you know, when you
13:15
have, when you have power come in, you want to have two
13:20
substation feeds and then you want to go to 2 medium voltage
13:23
houses and you're, and you're protecting yourself not only
13:26
just from like, like you said, you know, when you take
13:28
maintenance or, or there's some major retrofit, you need to be
13:31
able to isolate a portion of the building and then you need to be
13:33
able to tie it over. So we have like load ties where
13:35
it'll be like, all right, I'm going to feed the AB side of the
13:38
building from the CD side of the building or whatever that
13:40
configuration is. So there's that side where
13:43
there, there's a physical redundancy.
13:46
And the same goes for mechanical.
13:48
We do the same thing with having additional capacity for cooling.
13:51
We say the hottest day in Texas is going to be 100 and 30° and
13:56
that's only happened once in the history of time.
13:58
But like we're going to build to that design day because it gives
14:01
us redundancy. And then, but then there's also
14:04
like a logical redundancy. So you know, one of the things
14:07
that I've done, you know, being on the on the maintenance side
14:10
is, hey, I'm going to take down this entire switchboard and this
14:12
entire switchboard feeds sees 10 RP PS which are like reactive
14:17
power panel. So they feed the servers, which,
14:19
you know, basically the racks, they keep the racks online.
14:22
So when I take this thing down, all of this is going to go down,
14:24
OK? It's like everything downstream
14:26
is going to go offline. But now we have like a lot more
14:29
sophistication where we say, OK, we're going to alert service
14:31
owners. And this isn't always true,
14:33
Chrissy, if you like, think back to our first episode, like what
14:36
kinds of data centers, like this is not always the case.
14:38
This is where, you know, maybe a data center owner owns all their
14:41
traffic. If if you're dealing with like
14:43
infrastructure as a service where you have like many, many
14:46
owners, this is a lot more complicated.
14:48
So again, I'm oversimplifying, but you could say, all right,
14:52
we're going to shift all of the traffic to our sister site in
14:55
Utah now. So within a data center, you
14:58
always have capacity to take over what they call it like
15:01
basic. It's like a doomsday scenario
15:03
where it's like, I can now support all the traffic from
15:06
this other site. So all the servers have like 50%
15:09
capacity. So Nick, to your point, there's
15:12
like a lot that can happen to make this like both both
15:16
physically, logically and like from a process perspective,
15:20
create redundancy so that, you know, services and information
15:24
is like never lost or interrupted.
15:27
So I'm going to quote from one of my favorite movies, Data
15:31
Center is so hot right now. It's they've been a thing for a
15:35
while. But my goodness, the demand.
15:37
And you know, they're so popular, but there's a huge
15:40
amount of demand right now for data centers.
15:42
We're hearing about it everywhere.
15:44
Every organization, anything related AI is spinning up a data
15:47
center. What's what's driving that
15:51
demand? Why is it changing and what's
15:54
driving these trends? Yeah, I think the big the, I
15:58
mean, you know, the elephant in the room, so to speak, is, is
16:01
definitely this, you know, the the AI infrastructure and the
16:05
technology. So I mean it's it's Even so much
16:09
as like we have like the chat GPTS it like the large language
16:13
models and that are are kind of, you know, I'm sure you guys
16:16
noticed it's like integrated everywhere, right?
16:18
It's like in your zoom calls, I can transcribe this or in on
16:22
your phone now. It'll give you like the summary
16:24
of the text message and like, you know, it's kind of like
16:26
starting to embed itself everywhere.
16:28
And I think, you know, just the way that the way that technology
16:32
is like the next thing. And so, you know, just just same
16:36
way, like, you know, the short, what do they call like short
16:39
form videos, like the tik toks, the vines, the, you know, the
16:42
Instagram reels, all those things.
16:45
It's, it's the new way that we communicate and we relay
16:48
information is like in 5 to 8 minute videos or whatever.
16:52
The, I'm sure there's a sweet spot that, but I think that the
16:55
way that we exist and interact with one another is really
16:58
what's driving it. It's like this is what has
17:00
become effective. This is what, you know, we're
17:03
noticing engagement in these ways.
17:05
It's of course, you know, it's you can monetize on it.
17:10
Like I think there's all these different things and then, you
17:12
know, that's what what drives that.
17:15
Oh, we need a lot more infrastructure to support the
17:18
way that we're moving it around and interacting in the world.
17:21
Yeah, and it's. I was to say and, and it's
17:24
obvious when, Oh my gosh, something went down today,
17:28
right? Oh, it's down.
17:30
Oh, it's down. And it's, you know, the ripples
17:32
throughout the world even. And I think that kind of speaks
17:37
back to your, you know, your, your, your comments about
17:41
redundancy and how that is necessary, not just for the
17:46
building, you know, the one data center itself as itself on an
17:50
island, if you will, but in the data center Rome as a whole.
17:55
For sure, yeah. Yeah, the the the speed at which
17:59
I go to likedowndetector.com to see if the websites actually
18:02
down or if I'm just having a unique error on my it is so
18:07
funny. And then I really want to like
18:09
hammer on to like how AI and the traction that it's gone because
18:13
I've seen like such incredible laggards to adoption just go
18:17
completely all in two things. Like, yeah, not not to dog on my
18:22
dad. He's a wonderful person who's
18:24
taught me many things. He still watches his television
18:26
and standard definition because he doesn't want to press the
18:28
little red button on his remote control that brings it over to
18:31
HD. And for reasons I'll just never
18:34
understand. And that's totally fine.
18:35
I've made peace with it. But he has adopted Chachi PT.
18:38
He loves searching things on there, making images, all sorts
18:42
of other things. And it was fascinating to hear
18:45
like from Sam Altman when they were talking of the image
18:48
generator and when they launched it and how it just completely
18:52
changed the demand on their servers and like they were
18:56
literally melting their racks is at least what they were saying.
19:00
Just, you know, I, I think you hit the nail on the head.
19:03
It comes down to like how people communicate and how this has
19:06
become a tool for, you know, democratization that, you know,
19:11
you look at the art community and there's some folks who say
19:14
this is the worst thing that's ever happened because like, our
19:16
humans aren't going to make art anymore.
19:18
And then you have other folks who are like, well, this is the
19:20
best thing that's ever happened because now everyone can make
19:22
art. And at the end of the day, I'm
19:24
still an artist because my art's special.
19:26
And whether I use AI or whether I use a marker or a piece of
19:30
paint, and that's the medium I choose through to communicate.
19:34
And that gets back down to like what makes art beautiful And you
19:37
know, so many, it's that speaking to the human soul
19:40
aspect that you know, it's going to be human plus AI that makes
19:44
great high quality art. It's not going to be AI by
19:48
itself because AI is a partner to us.
19:50
We still have to prompt it. It's not just generating art.
19:54
On the way, you know, on the side, unless it is, I always
19:59
think it's such a weird like a weird take and I I understand.
20:02
I'm not understand. Start off by saying I
20:04
understand. But when somebody says like
20:06
almost AI wrote that, I always think if if it wrote it without
20:11
a proc, though, like that would be one thing, but there's
20:13
obviously enough context there that someone you know, dumped
20:16
their, you know, their thoughts into it.
20:18
And so I like to actually take a very different stance.
20:21
I I see like you know. Train it, make it more you, you
20:25
don't want to sound robotic because it's like it takes away
20:27
some of that human element. But like, I don't want to spend
20:31
an entire afternoon writing, you know, an e-mail update for a
20:36
project if I can do it in 5 minutes.
20:38
And like, I don't know, I, I just like have a different, you
20:40
know, different take. Aren't like, yeah, use it, be
20:43
productive, save yourself some time.
20:45
Absolutely. We very much embrace the work
20:47
smarter, not harder. And you know, I've been hearing
20:50
that for a really long time in my career to the point where
20:54
like, OK, I can't literally go out on a on a shop forum and
20:58
build these things faster. So cool, thanks for telling me
21:01
to work smarter, not harder. But this was one of this has
21:05
actually been like, this is working smarter, you know,
21:08
taking the tedious and redundant, you know, and making
21:12
it so much easier. Yeah, and I like it too.
21:15
It just, you know, sometimes I don't work good and sometimes I
21:18
work good and then I trust myself.
21:20
But sometimes I don't work good and or, or many times I can be
21:23
pretty direct and there's a lot of times where I don't,
21:27
especially because we're like in the world of Slack
21:29
communication. I know I read between lines.
21:33
I'm like, can you take the whatever I was feeling when I
21:36
wrote set out of here and just like make this a constructive
21:40
message so no one knows that I'm like walking around with an AXA
21:43
grind and I'm really actually happy with what was done.
21:46
And I, these couple things need to change and we're going to be
21:50
like, we're better than we've ever been.
21:51
And this is really awesome. And it does such a good job of
21:54
like helping me find my words or my intention, which is why I
22:00
tend to trust it more. And I tend to like having it
22:04
right for me. And maybe it's not my quote
22:06
voice, it's my now AI voice. But you know, I've read the
22:10
thing, I prompted the thing, I knew what I wanted to change and
22:13
AI helped me get there. And I think that that's like so,
22:15
so, so important. And we do.
22:17
It did go off on a little bit of a tangent here because I think,
22:20
you know, AI is super cool. We're all using it.
22:22
But I would love to kind of go back to the failure to means
22:26
just a little bit more because I think there's something that's
22:29
really, really unique about this is how do modern data center
22:34
teams build smarter and not just more?
22:38
And I think what I mean by that is like redundancy often times
22:41
means building 10 to Power 8. And how do you find that number,
22:46
that sweet spot so that you're not just building more your
22:51
understanding, I guess your failure tolerances, if you will,
22:55
and how to accommodate that? Yeah.
22:57
And, and this isn't, this is like a very highly debated topic
23:03
in the industry right now because.
23:04
I know you would have the answer because it's so nebulous, but
23:08
I'm curious. Yeah, yeah, yeah.
23:09
Well, and and it's, you know, it's kind of driven by some of
23:12
these big, you know, chip manufacturers.
23:15
I'm not going to name names, but there, there's several of them
23:18
that are, you know, providing direction.
23:21
Is it the right direction? I don't know.
23:24
I, I don't know, I don't think anybody's done it to say like,
23:26
yes, this is better than like, you know, your example, like,
23:30
you know, we, we say, you called, you said it in the other
23:32
episode. We have 3 nines.
23:33
We have 5 nines. And then we have things like 5
23:36
to make 4 three, you know, four to make 3.
23:38
And that can be, you know, and there's different ways you can
23:41
do that too, like 4 to make three could be you have one.
23:44
This is what I've in my experience at data center, just
23:47
generally what we've had is we've been a basically a, a
23:50
board that doesn't have anything on it.
23:52
It doesn't carry any load. And then if you have another
23:55
board fail, it will pick up the load.
23:57
So it's, it's like a, they call it a catcher or a reserve board.
24:01
Well, now we're looking at it differently.
24:02
We say, OK, no, we can actually have four boards, but they're
24:05
all loaded at 80% instead of 100%.
24:08
So that if one fails, we can distribute that we have a
24:11
distributed redundancy instead of a block redundancy.
24:14
So there's like a lot of different ways that, you know,
24:17
we're, we're evaluating this like redundancy model in the
24:19
data center industry right now. And again, I don't know the
24:23
right answer and, and there's a lot of opinions on it.
24:27
But, but I think the, the big thing is like, how do we, you
24:32
know, how do we also like build in redundancy in, you know, down
24:37
to like the chip level or, you know, how do I think that, you
24:41
know, the distributor redundancy model kind of speaks to what
24:43
you're saying, Nick, It's like, OK, well, maybe instead of just
24:45
having equipment sitting there unused, maybe we just load it
24:49
differently. So I think there's a lot of
24:50
different ways that it can be done and, and there's a lot
24:53
smarter ways that we can, we can build and deploy and that that
24:57
goes into a different episode. But you know, thinking about
24:59
like prefabricated, it's skidded equipment utilizing like AI and
25:04
technology and building like the, the cool little robot dogs
25:07
and all these other things. So beyond just like how we
25:10
operate them, I think there's a lot of ways that we can explore,
25:13
like how to be smarter, which I think perfectly ties up this
25:17
entire episode really is like we have a limited workforce.
25:20
And I think that's something that we will continue to
25:23
struggle with because these buildings are bigger, they need
25:26
more people. And we're not seeing a lot of
25:28
people go into trades right now, not like we, we have in the
25:31
past. And so how do we deploy
25:34
technology that, you know, I, I saw on the other day, look this
25:36
up, it's really cool. But it's like the little robot
25:38
that puts the shingles on the roof, you know, and it's maps it
25:42
all out and then it, it lays the shingle and it nails it in and
25:45
then it moves and it pivots. And it's like, how do we find
25:48
more opportunities like that to like reduce the the hard manual
25:51
laborers that people aren't burning themselves out, but we
25:54
still get people in the industry.
25:56
Absolutely. And that's I think we could go
25:58
at least a couple more episodes just talking about about that
26:02
topic. I know Nick and I have have
26:03
touched on it before. I'm going to shift topics.
26:07
This may put you in top five. This could be the moment right
26:10
here. OK, I'm ready.
26:11
You have an amazing reputation for your knowledge and data
26:16
centers, the work you're doing with our customers, just your
26:20
smarts and your savvy and your communication.
26:23
You are also really well known as the class clown in all of our
26:27
team happy hours. And there's never a meeting that
26:30
goes by that. I don't look for Kaylin
26:32
witticism in something that has happened.
26:36
So I know you're exploring comedy and tying that into
26:40
communication. What what took you there and and
26:43
how do you see that unfolding for you?
26:46
Yeah, that is my proudest title, 2024 Avocado Class clown.
26:54
I don't expect anybody else to carry the title, so it is a lot
26:56
of weight, but I. Heavy is the head.
26:59
Heavy is the head. I mean, tell you, it really is.
27:02
I wear my sash to bed at night so that my husband knows who's
27:07
the funny one. Actually, after, after the pit
27:10
stop where I was, was crowned the class clown, I, I'd had a
27:15
several conversations and people were saying, what are you going
27:17
to do now? Like, what do you mean?
27:19
What am I going to do now? I won $200.
27:20
Are you kidding me? Like this is my, my first paid
27:23
gig. I'm done.
27:25
And so, but it, but what it actually did, it was like, it
27:29
made me go, you know, what could I do with this?
27:31
Like what, how can I lean into this more?
27:33
And, and so I actually started doing improv classes after that.
27:37
And, and improv is interesting because it forces you to
27:43
communicate on a whim because you don't know what the other
27:47
person is going to do. It's like kind of the beauty of
27:49
it is like you can be as prepared as you want to be, but
27:52
it doesn't really matter because you don't really know where
27:55
things are going to go. And so this, this has been like
27:58
kind of my journey this last year has like 1 doing the improv
28:03
classes and going to like open mic nights and it, and it's
28:07
helped me a lot to one, not only like, you know, there's
28:11
something about public speaking that is required.
28:15
You're in front of a crowd of people and that's helped me a
28:17
lot with my communication. And two, you know, you have this
28:21
common ground with people at these shows, like everybody's
28:23
there to see a comedy show or they're there for the class or
28:26
they're doing their first open mic.
28:27
And so it creates this common ground.
28:29
And so even that can be a little awkward.
28:31
I'm sure you guys experience it like at conferences, you know,
28:34
where somebody comes up and it's like, how do I find the thing
28:36
that I can talk to you about? Like there's got to be this
28:38
thing. And so it's helped me like
28:40
identify the, the common ground and then and then beyond that,
28:47
it is always a really good way to like break up tension.
28:51
I mean, we're in meetings with like suits and ties.
28:53
And I have had the pleasure of like, you know, absolutely.
28:59
I mean, Nick's been there. It's like, all right, how can I
29:02
break this up? You know, like come in with a
29:05
good joke or like, you know, with a witty come back or
29:08
something. And it just changes the entire
29:11
like landscape. Yeah, I have never thought that
29:14
I'd go to a data center planning meeting and talk about hot dogs
29:17
and hamburgers for as much as we did and jalapenos too.
29:23
I have to ask every every different vegetable or, or food
29:27
item that you come up with and what it actually means in in, in
29:30
data center lingo. So really, you know, huge thank
29:34
you for being on the show. We always like to wrap it up
29:37
with one final question. So you know what's something our
29:41
audience, our listeners should be reading, trying or
29:43
experimenting with to stay wired to build?
29:46
Oh, OK. Because I've mentioned it a
29:49
couple times. I'm like obsessed with all of
29:51
these little construction robots.
29:54
Like it has like taken over my algorithm.
29:58
And so I would encourage people to what like what I mean, maybe
30:04
this is like a base level, like look into these things, like
30:06
just understand what's happening out there.
30:10
And, and then beyond that, like check out like some robotic
30:15
conferences. Like there is like a unique tie
30:17
in to data centers. We go to the data centers and
30:19
there's a lot, a lot of like suppliers, like vendors there
30:22
that have like, you know, some critical equipment.
30:25
But like the robotic conferences seem like they've got some
30:28
really cool stuff was like cutting edge and like, how do we
30:32
build smarter, not harder? And so I, I would encourage
30:36
people to, to dig into that is just like a, it's, it's a fun
30:41
way. It's easy to stay engaged.
30:42
It's not as boring as like looking up ISO standards for
30:44
construction. So, you know, maybe start there
30:47
and then and see where it kind of kind of leads you.
30:50
Yeah, and these robots need operators too.
30:53
So if you like flying drones or whatever, maybe that's your
30:55
point of entry into this phase. So and.
30:57
I know. Thank you again.
30:58
It's always awesome. Yeah, right.
31:01
Yeah. Have a just like a little fleet
31:03
of robo dogs. But what do you give them for
31:05
treats? I have no idea.
31:07
Still kind of scary. Sorry, guys, they're still kind
31:10
of scary. Oh my gosh.
31:13
They're because they're not fair and they got to add fur that
31:16
way. They we.
31:16
Know that they're I've just seen seen too many sci-fi dystopian
31:19
movies. I think so.
31:21
Yeah, that's what it is. Oh my.
31:24
Goodness. Well, thank you so much.
31:25
Always a pleasure. I will see you next week at the
31:29
in Bend and then Orlando. So we got a loaded, loaded week
31:32
coming up. So excited to have this
31:34
conversation ahead of then and I'll see you soon.
31:37
Thanks for being on the show. All right.
31:38
Sounds good. Bye guys.
31:40
Thanks, Caitlin. Wonderful episode, always enjoy
31:45
getting to bring someone else from the team on.
31:47
And So what did you think Kirsty?
31:49
I know this is probably your first stab at learning a bit
31:53
more of what Caitlin is up to, I think.
31:55
More of the current stuff that she's up to.
31:57
I mean, I knew her background. I knew her, her, her voyage, if
32:00
you will, from the Coast Guard to data centers and what that
32:04
looked like and that she's always been her class clown at
32:08
heart. And to get the actual award was
32:10
pretty great too. But just understanding what she
32:13
does in the industry and how she helps make those connections is
32:17
awesome and, and actually surprises me.
32:20
Not at all. She's so dynamic and and
32:24
wonderful to be around. And yeah, I'm a key member.
32:29
I'm going to be a founding member, but I am definitely a a
32:32
high level member of the Kaylin Bushel thing called.
32:36
Yeah, we're gonna have to get that e-mail box set up.
32:39
She does. It's just it's wonderful
32:40
because, you know, I think we mentioned class kind of like 100
32:43
times on this episode. And there's like a part of my
32:45
brain that was like, is this something we should be
32:48
celebrating? There's always that, like,
32:49
should we be celebrating the goofing around?
32:52
But I think there's the the reason Caitlin has won that
32:54
award, at least from my perspective, is just, you know,
32:57
to, to win an award like that and to then continue to be well
33:00
respected is I think a testament to her fearlessness within the
33:05
space to be able to speak up, to ask the dumb question.
33:08
Do not be afraid of sounding stupid or silly, because at the
33:12
end of the day, we're, we're all probably feeling that way a
33:15
little bit. And she acts with the the
33:17
bravery and the tenacity to add value or add that question so
33:23
that we can have the communication.
33:24
And I think, you know, I wonder to any of our listeners, if you
33:28
do kind of superlatives at your company and what they might be
33:32
if you have a class clown and they're a value add or maybe a
33:35
value detractor for some of us. But what those, what those are,
33:40
if you want to share them with us on one of our posts are on
33:44
our on our channel, let us know. But you know, I think those are
33:48
the things that one thing that I know that Avocado has really
33:51
taught me since my 3 1/2 years of being here is that creating
33:55
room for that stuff is what creates room for the human
33:58
connections. And you know, when I reflect on
34:01
construction and how brash we can sometimes be, it's largely
34:05
because we're so focused on working and not the and building
34:08
the teams that get the work done.
34:10
And I think that that's something that's really special
34:12
about Avocado and special about the companies that we work with
34:16
is creating that shared culture to be able to be vulnerable, to
34:20
ask the dumb questions and then to work together to contribute
34:24
towards the best answer possible at that time.
34:27
So that's what I'm going. To be thinking about, yeah.
34:30
Absolutely. Our one of our key tenants said
34:32
be your authentic self and you cannot tell me that that is not
34:35
a Caitlin's authentic self. So I have a big appreciation for
34:39
it. Another thing that struck me is
34:42
just that experience she mentioned about, hey, we're, you
34:46
know, we're sticking a fan in a plenum and cool, that job's
34:50
done. But nobody thought about what
34:53
happens next or when it needs to be serviced or fan breaks or
34:57
something needs to happen. And you know, it really speaks
35:01
to the kind of asset, asset centric focus of the building
35:07
and having to look at what you're doing holistically and
35:10
the value that that brings from the beginning, right?
35:13
Starting with the end in mind, understanding what that journey
35:16
looks like, not just from the beginning phases of
35:18
construction, but all the way through operation and
35:20
maintenance is so valuable. There was probably a joke in
35:25
there she could have written of how many people does it take to
35:28
change a fan belt in the thing in a plenum.
35:30
But you know, we'll leave that where it is.
35:33
But I really loved that that example of you can't, you know,
35:38
having, having people who are able to lift up their head and
35:41
look out beyond just the job at hand or the task at hand.
35:44
So I think you know, with that we're going to wrap today's
35:47
episode. Thank you so much for listening
35:49
in. If you are going to be at 7 by
35:51
24, I know you're listening to this after we hope we ran into
35:55
you during that. We're actually going into our
35:57
slowest travel season. So I don't think we have much of
36:00
anything for July or August in terms of being on the road.
36:03
You will if you are at Infra Day Midwest up in Chicago, running
36:07
to us there. But if not, enjoy your summers.
36:10
We're still going to be recording some episodes.
36:11
We got tons of fun content planned for the rest of the year
36:14
and looking forward to continuing the stay wired to
36:17
build with you all. So with that, have a great rest
36:20
of your month and we'll see you on the next episode.