0:00

I mean, first off, any palletier person

0:02

who's listening to this, we don't

0:04

really use the word centers of

0:06

excellence inside Palantir. I mean, there's

0:08

going to be functions of people

0:10

who need to be experts and

0:12

things, right? That's just de facto

0:14

expertise exists and it has value.

0:16

But grouping those folks far away

0:18

from where the work is getting

0:20

done or where it's needed is

0:22

a pretty dangerous proposition, right? Welcome

0:24

to Technivation. I'm your host Peter

0:26

High. Jim is the chief information

0:28

officer of Palantir, a technology firm

0:31

that connects data, analytics, and operations,

0:33

allowing users to analyze data and

0:35

take action, that earns in excess of

0:37

$3.7 billion in annual revenue. Jim has been

0:40

with the company for a dozen years, three

0:42

of them is CIO. Jim leads Technology

0:44

for a firm that continues to grow at more

0:46

than 30% annually, and he does so with a

0:48

team that's remarkably efficient. This team is

0:51

also customer zero for the company

0:53

and its products. His journey is all

0:55

the more remarkable remarkable. when you learn

0:57

that up until he was 30,

0:59

his life was focused on music.

1:01

I look forward to hearing more

1:04

about his remarkable journey through this

1:06

conversation. Jim, welcome to Technovation. It's

1:08

great to speak with you today. Thanks so

1:10

much, Peter. It's good to be here.

1:12

Wonderful. Well, Jim, I thought we'd begin

1:14

with your company itself. Palantier, talk a

1:16

bit about what it is that the

1:19

company does, if you would. Many years ago,

1:21

my boss or CTO, Sean Sankar,

1:23

was giving a TED Talk, and

1:25

the way he framed it, and

1:27

it's long before I worked at

1:29

the company, I remember watching this

1:31

TED Talk, he said that it

1:33

was about, Palatir was about letting

1:35

computers do what they do best,

1:37

so that humans, analysts in those

1:39

days in the military and the intelligence

1:42

world, could do what they do

1:44

best. So that's, it's evolved into,

1:46

it's the, Optimizing the

1:48

union of the computer and

1:50

the human doing whatever the

1:53

work task is that they're

1:55

doing. So there's a big

1:57

emphasis on decisioning on the

1:59

analysis. whatever pile of data a

2:01

company or an agency or a government

2:04

has a yielding a solution to whatever

2:06

the biggest problems are that they've got.

2:08

And that can take on a bunch

2:10

of different forms. And the nature

2:12

of your work and the fact that

2:15

so much of it actually is with

2:17

governments means a lot of what you

2:19

do is secret naturally and as a

2:21

result there's a lot of mystery around

2:23

Palantir's business as well. It's something it's

2:25

in some way been kind of like

2:27

a beneficial calling card of sorts, I

2:29

suppose, it makes it more magnetic in

2:31

some ways. I wonder what you thought

2:33

is about that. Well, you know, in some

2:36

ways, yes, in some ways, no. I

2:38

mean, I think it's just a, I

2:40

think it's just a truth that whether

2:42

it's, whether there's actually classif, like data

2:44

classification laws or rules, or whether it's

2:46

just, you know, even for a company

2:48

that it's not in the most important

2:50

problems, that they have the least idea

2:52

how to solve. are the things they

2:54

feel most sensitive about. And the solution

2:57

to those things ends up being part

2:59

of their competitive advantage. So we're pretty

3:01

accustomed to dealing with these differing levels

3:03

of, you know, we're always grateful when

3:05

we can talk about the stories. Sometimes

3:07

the really cool ones are the ones you

3:09

can't talk about, but yeah, we find a way

3:11

to get the word out as much as we

3:14

can. Try to. Talk a bit about your

3:16

role, if you would, as Chief Information, if

3:18

you would, as Chief Information Officer, Can you

3:20

talk a bit about how you fit

3:22

into the broader organization? Yeah, there's this,

3:24

there's a running gag that, you know,

3:27

the only reason that we have a

3:29

CIO Appellant here is so that somebody

3:31

goes to jail if we do the

3:33

wrong thing. We're very comforting to hear

3:35

that. I know, it's extremely comforting. Yeah,

3:37

I question my decisions over and over.

3:40

We're a company that lets... IT organizations

3:42

and technical people all over companies kind

3:44

of set free to do their work,

3:46

right, which we can talk about throughout

3:48

our conversation today. So it puts a

3:50

role like Cio in this really interesting

3:52

light. I mean, and look, like the

3:54

Cio as a role is different in

3:56

basically every company anyway, right? Like there's

3:58

no two that are alike. really, at

4:00

least the ones that I talked to.

4:02

In my world, the CIO role

4:04

is very internally focused and

4:07

very operational. And that's actually

4:09

true with pretty much every

4:11

really responsible job or leadership

4:13

job at Palantir. We're a

4:15

company that doesn't tend to

4:18

delegate operational things away. We

4:20

tend not to delegate execution

4:22

away. So. Most of my

4:24

time 75% plus of my

4:27

time is actually running IT

4:29

and aligned adjacent things within

4:31

Palantir. Among those things,

4:34

all of our internal cloud

4:36

rolls up to me, all

4:38

of our compute, storage,

4:40

infrastructure, security systems, some

4:43

off, obviously IT

4:45

operations, technical compliance.

4:48

a lot of our business systems and financial

4:50

systems and that sort of thing, bookkeeping

4:53

systems. So, and adjacent to that, there's

4:55

also, I have some product responsibilities,

4:58

we have several dev organizations within IT,

5:00

they're building various different things, which again,

5:02

we can also talk about as we

5:04

go through the conversation. So. Yeah and then

5:07

with the rest of my time I'm I'm

5:09

having conversations like this or working increasingly with

5:11

customers who are interested in how we do

5:13

what we do and how you know how

5:16

our execution leads us to have the products

5:18

that we have and how they can can

5:20

learn from very interesting I was

5:22

fascinating in from our past

5:24

conversations to learn how efficient your team

5:27

is and one that has become more

5:29

efficient across time yeah one would think

5:31

I'd with the remarkable growth through now

5:33

you know closing in our four billion

5:35

dollars and annual revenue that you must

5:37

have a massive organization around you and

5:39

that's not the case. Talk about why

5:41

that isn't the case and how you can

5:43

do what you've just described within the

5:45

grand scheme of things, relatively few people.

5:47

Yeah, yeah, there's this whole, there's a

5:49

big conversation now with a lot of

5:51

our customers and prospective customers about like

5:54

you can run on Palantir or you

5:56

can run like Palantir and those two

5:58

things make most sense when there together.

6:00

So this is a run-like Palantir conversation.

6:02

You know, when I, I've been in

6:04

the, in the org or around IT

6:07

org in Palantir 12 years, like you

6:09

mentioned, and you never really think when

6:11

you're kind of growing up in an organization

6:14

and doing all these different jobs

6:16

throughout it, that one of the things

6:18

you're going to end up doing is

6:20

dismantling it, right? Like you never really

6:22

seek that out. That's not, you assume

6:24

you're going to be building it

6:27

further. When I took over

6:29

a CIO, the business was

6:31

changing really drastically, right? Our

6:33

commercial product, Palantir Foundry, was rapidly

6:35

maturing. More and more companies

6:37

were using it to solve their

6:39

problems, and they were getting a

6:41

lot of efficiencies out of

6:43

it. I looked at how we had built the

6:45

IT organization here, and it was big.

6:48

It was, you know, with with full-time

6:50

internal employees and contractors and

6:52

consultants, it was over 200

6:55

people at that time. And

6:57

there's this, there's this cultural

7:00

value inside Palantira that's it's

7:02

about like trying to decalcify something

7:04

like the more something is an institution

7:06

that must be propped up and nobody

7:09

can imagine life without it kind of

7:11

the more your instinct should be to

7:13

try to break it. Right. So actually

7:15

I read a I read a Wall

7:17

Street Journal article. This is a long

7:19

time ago by a guy named Joe

7:21

Pephart, who is an MIT professor, business

7:23

professor at MIT at the time, I

7:25

think he's at UC Davis now, called

7:27

it's time to abolish the IT organization,

7:29

something like that. They made the case

7:31

that putting the org in its own

7:34

bubble pretty much guarantees that it's always

7:36

going to be a daylight and a

7:38

dollar short compared to the business that

7:40

you're actually trying to power. It puts

7:42

you in a position of servitude, it

7:45

puts you in a position of like

7:47

you're a snack bar operator taking short-ordered

7:49

cook, sort of orders over-the-counter. And there's

7:51

just that puts an upper-down on how

7:54

good you can get. So we embarked

7:56

on this factoring out, federating

7:58

the organization. into the rest

8:01

of the company, actually making

8:03

some teams pretty substantially smaller

8:05

in the process just by

8:07

virtue of taking a hard

8:09

look at how we're grouped around the

8:11

work and what's actually needed in

8:13

the modern era to do that

8:15

work. Now what we've always thought

8:17

is needed. And over that time

8:19

we went from 200 or a little

8:22

over 200 to now our full time

8:24

our internal employee account in in IT

8:26

here is below 80. I think it's 70.

8:28

seven or something right now I should know

8:30

that number. And of course we still have

8:32

some some contractors and consultants but that

8:35

workforce has been greatly reduced as well

8:37

because it just makes more sense to

8:39

us now the way we operate the

8:41

in-house a lot of those functions. It's

8:43

been a pretty interesting journey not something

8:45

I ever would have thought that I

8:48

would have actually pushed out let alone been

8:50

part of. Yeah it's remarkable and and don't

8:52

need to tell you how different it

8:54

is from most of your peers. And

8:56

you've noted and again in our past

8:58

conversations that there's some some reasons that

9:00

this works You stand to allude to

9:03

some of them, but I wanted to

9:05

highlight four points you at least highlighted

9:07

to me and then cover each of

9:09

them with you That you mentioned the

9:12

pivot to customer zero a democratized delivery

9:14

and development a different delivery model and

9:16

then no centers of excellence Let's cover

9:18

each of those intern if you don't

9:20

mind beginning with customer zero. Absolutely, and

9:23

I'll just to be up front these are

9:25

all facets of the same thing. These

9:27

are all ways that we've been trying

9:29

to spin a certain type of flywheel

9:31

faster and faster. So as I'm telling

9:33

this story, I'll try to link them

9:35

all together. So customer zero, you know,

9:37

are like any company you want

9:39

to run on yourself, right? Palatier's been

9:42

around for 20 plus years. Palantier Foundry,

9:44

especially Foundry with AIP, as it is

9:46

now in its modern iteration, much, much

9:49

less than that. So we had a

9:51

similar problem to a lot or similar...

9:53

opportunity, I guess, to a lot

9:55

of our customers, which is we were working off

9:57

of a 5 to 15 year old tech stack.

10:00

and set of processes and human structures

10:02

as we talked about, and looking at

10:04

something that was vastly better. So in

10:06

late 2023, we kicked off a series

10:09

of change initiatives to get to the

10:11

point where we are now where we're

10:13

aggressively adopting foundry as a replacement product

10:16

for a lot of commercially sourced things.

10:18

and trying to do that in a

10:20

way that models for customers as well

10:23

and feeds back to the to the

10:25

product organization. Very interesting. Talk a bit

10:27

about this democratized delivery development that you've

10:30

also highlighted. Sure. Yeah, one of the

10:32

feature, democratized development in delivery is one

10:34

of the features. One of the main

10:37

calling cards of Palantir Foundry, actually. It

10:39

has to do with how the security

10:41

construct. the object layer of the data,

10:43

which we call the ontology, like a

10:46

semantic object set that represents all the

10:48

nouns in your organization. How you can,

10:50

if you permission that correctly in the

10:53

way the products design, you can set

10:55

free a lot of people towards the

10:57

edges of your organization to build whatever

11:00

they need to build to solve their

11:02

problems. Of course, you have to have

11:04

a governance construct and all of these

11:07

other things kind of set up to

11:09

deal with that. So you don't just

11:11

unleash the bad kind of chaos in

11:14

your organization. But what we found is

11:16

it unleashes the good kind of chaos

11:18

in the sense that instead of letting

11:20

roadmaps like long-term roadmaps be the defining

11:23

thing about change in IT but also

11:25

in the rest of our internal functions

11:27

we found it kind of flipped the

11:30

script where all of a sudden people

11:32

at the edge of the organization were

11:34

frustrated with process or frustrated with the

11:37

tools we had given them. We're starting

11:39

to but it's like it's like you're

11:41

legitimizing shadow IT. You're asking people to

11:44

pull. IT in the direction that their

11:46

work actually needs it to go, which

11:48

was intrinsic to our adoption of foundry,

11:51

and I don't think we could have

11:53

really done one without the other. Very

11:55

interesting. And then you talked about a

11:57

different delivery model was a key catalyst

12:00

as well. Talk more about that if

12:02

you would. Yeah, I mean, this is

12:04

kind of, this gets into the governance

12:07

story a little bit as well, right?

12:09

There was always a, and this kind

12:11

of pre- games, the Centers of Excellence

12:14

Conversation too, you know, that we always

12:16

had a data engineering team, for example,

12:18

or an internal dev team who served

12:21

the needs of all the functions. Palants

12:23

here, right? And that's just in 2014-15

12:25

when those things came about. That's actually

12:28

fairly forward thinking, right? And it worked

12:30

and it was great. What we found

12:32

was that as the scale of data

12:35

was growing, there's actually no way to

12:37

build an internal, especially if you're trying

12:39

to make the orch smaller and federate

12:41

it within the company, there's no way

12:44

to build an internal data engineering organization

12:46

that is close enough to the actual

12:48

source of the use of the data.

12:51

to have any understanding of what's actually

12:53

going on. And so what we found

12:55

was as a scale that data was

12:58

growing, our rate of, you know, spilling

13:00

data on ourselves or needing to remediate

13:02

things or reliability or accuracy data freshness

13:05

issues was outpacing the rate of growth

13:07

of the data. So we had to

13:09

think about how do we deliver the

13:12

data to the edges of the organization

13:14

where these builders are. in such a

13:16

way that they can actually do the

13:18

right thing and stay on the right

13:21

side of standards compliance and have fresh

13:23

and accurate data. So one of the

13:25

things we've been working on is having,

13:28

instead of a data engineering data governance

13:30

group, instead of having all the systems

13:32

engineers across all the rest of IT

13:35

and adjacent teams, including the HR team,

13:37

legal, some others, procurement team, any team

13:39

that owns or operates a system that

13:42

sources data. into the warehouse, having those

13:44

folks be the ones that are responsible

13:46

for the compliance correct. freshness performance of

13:49

the data, not just up to the

13:51

boundary of Palatier Foundry, but into Foundry.

13:53

So that requires a bunch of retraining

13:55

and a bunch of upskilling for those

13:58

folks. But what it means is if

14:00

I'm a person at the edge of

14:02

the organ, I need access to HR

14:05

data, the person who's going to grant

14:07

that to me is actually somebody in

14:09

HR. And if I have a problem

14:12

with that data. The source of that

14:14

data in HR, workday in our cases

14:16

are each CM, the people who run

14:19

workday, both the HR folks and the

14:21

IT folks, are going to be the

14:23

people answering my data ticket as well,

14:26

not the data engineering organization. That change

14:28

and delivery mashed with that changing governance

14:30

unlocked a huge amount of velocity for

14:33

us. Obviously I could talk about it

14:35

forever too. I wanted to get into

14:37

the... the last of the four which

14:39

is no centers of excellence which is

14:42

such an interesting just just hearing the

14:44

words is such an interesting concept peel

14:46

back that I need it be will

14:49

yeah yeah this is actually this is

14:51

linked to one of the big things

14:53

I'm trying to get done this year

14:56

I'm trying to get done this year

14:58

at IT too is the kind of

15:00

like reshaping the organ rethinking how we

15:03

group people around the outcomes you know

15:05

the centers of excellence thing maybe first

15:07

off any palletier person who's listening to

15:10

There's going to be functions of people

15:12

who need to be experts and things,

15:14

right? That's just de facto expertise exists

15:16

and it has value. But grouping those

15:19

folks far away from where the work

15:21

is getting done or where it's needed

15:23

is a pretty dangerous proposition, right? Like

15:26

it's that snack bar mentality that I

15:28

talked about a couple of items ago.

15:30

If you, as a person who provides

15:33

that expertise or somebody who runs that

15:35

center of excellence, you want to you

15:37

want to create focus for your people,

15:40

but what it actually does is it

15:42

it's a layer of separation from the

15:44

actual work getting done. So we're always

15:47

always looking for ways to break a

15:49

center of excellence and put it closer

15:51

to the business. Like we're in the

15:53

process of our product organizations and the

15:56

process of doing that with part of

15:58

recruiting right now, right? Where it's like

16:00

to get the recruiting organization closest to

16:03

the places where the recruitment actually needs

16:05

to happen, you actually have to put

16:07

them right next to each other. So

16:10

they're all in the bow growing

16:12

together. It's kind of like, it's not

16:14

a new idea, but for us, like

16:16

most things in Palantir, we lean into

16:18

it in this really hardcore way and

16:20

just kind of, and try to work

16:22

through it and try to iterate quickly

16:24

and get it done. So that's how it's,

16:27

that's how it's being in festivals. cuts

16:29

against the grain of a lot of

16:31

your peers as CIOs. And I wonder

16:33

at me to what extent, obviously yours

16:35

is a very different company, but

16:37

also yours personally is a very

16:39

different background. You don't come with

16:41

any baggage of having been a CIO

16:43

somewhere else or having any role in IT

16:45

anywhere else as you grew up in IT

16:48

in Palants here. We'll get a little bit

16:50

further into the specifics of the early stages

16:52

a little later, but you know, I wonder

16:54

the extent to which that's been

16:56

helpful to you. That's been helpful to

16:58

you. to have to not have inherited

17:01

wisdom of what must be based upon

17:03

how most other peers operate their IT

17:05

shops. Is that fair? I think that's

17:07

fair and it's a good way to

17:10

say it. If you look at most

17:12

people who hold a lot of

17:14

responsibility at Palantir, I'm not going

17:16

to say I hold a lot,

17:18

but I hold some. Most of us

17:20

grew up in the company, to

17:23

an extent. There are notable exceptions,

17:25

obviously, but I think by

17:27

and large, internal folks make up

17:29

most of the folks who carry a

17:31

bunch of responsibility here. And it's

17:34

for that exact reason, right? Like

17:36

if you want to, if the whole thing

17:38

is about. defining a type of Orthodoxy, certainly

17:40

an IT that's the case for me for

17:42

what I'm trying to build. But the company

17:44

in general, it's about running against a type

17:47

of Orthodoxy in thinking, right? And certainly our

17:49

customers value that. We try to generate Alpha

17:51

for our customers, right, and come with new

17:53

ideas that are breakaway, you know, in some

17:55

direction or another that generates whatever the outcome

17:57

is we're looking for. So it's hard to

17:59

prove. that kind of alpha, whether with

18:01

the customers or internally, if you do

18:04

things the same way everybody else doesn't.

18:06

So for a long time, it actually was a

18:08

source of imposter syndrome for me to

18:10

be honest. Like I would go to

18:13

these CIO gatherings and like not understand

18:15

any of the jargon that was happening.

18:17

I feel like I hadn't hopped the

18:19

same seven companies as everybody else and

18:21

I didn't play golf with these folks

18:23

and it's like, oh man, like, what

18:25

are we even doing in this room?

18:27

Now that the results are showing,

18:30

there's different ways to do things that

18:32

really can add to the conversation, right?

18:34

And that's something that it's a big

18:36

value, a palantier, that's a thing we

18:39

try to put out there in everything

18:41

that we do. Yeah, it's really

18:43

fascinating. And why do you suppose people, you

18:45

know, there's the performance of the

18:47

organization, or there may be obvious

18:50

reasons, are there any non-obvious reasons

18:52

as to why there are people

18:54

of such a long tenure like

18:56

yourself? who've chosen to stay for

18:58

so long. Operating, by the way,

19:00

in Amelia in Silicon Valley, where

19:02

it is duregur to hop from

19:04

company to company, I mean, you

19:07

have worlds of opportunities, you know,

19:09

within blocks of where you are

19:11

at all times, you know, it's

19:13

pretty extraordinary that that's the case. Yeah,

19:15

one long-time colleague of mine jokes, it's

19:17

because we've all gone feral, you know,

19:19

like there's not a, you feel like

19:22

we could survive anymore. For me

19:24

personally, I can answer that by saying, you

19:26

know, I've been CIO now, I'm in my

19:28

third full year, I suppose, and that's by

19:31

far the longest I've ever done anyone here.

19:33

And, you know, I think most people

19:35

who are at Palantir self-select to

19:37

be here because they're curious,

19:40

they're generative, like they're kind

19:42

of unsatisfied with sitting still.

19:45

Often they're overachievers, they're

19:47

overqualified for whatever they're doing, like these

19:49

really switched on people. And for a

19:51

lot of us, you've stuck around for

19:54

a long time, it's that constant new

19:56

set of challenges, or there's

19:58

this constant, like your... I'm always the

20:00

dumbest person in every room that I'm in

20:03

and I find that just addictive. Like there's

20:05

always some interesting thing going on. I feel

20:07

like I'm answering like an interviewee's question

20:09

like somebody who's applying to the company

20:11

going like why are you here but

20:13

it's actually true it's not just a

20:16

company line and I think a lot

20:18

of people who've been here more than

20:20

a decade would say exactly the same

20:22

thing. It's just like it's the right

20:24

kind of perpetually challenging, productive, productive discomfort,

20:26

and it's addictive. I want to talk a

20:28

bit about the year ahead. You shared with

20:31

me that you've got three different areas, your

20:33

focus from that perspective, continuing the

20:35

ongoing evolution of the customer's euro

20:37

program, reshaping the team to a

20:39

greater extent, and introducing greater levels

20:41

of resiliency and hardening the organization's

20:43

systems. I'd love to hear a

20:45

little bit more about the future

20:47

of each of those. We've talked

20:49

a bit about customer zero. And

20:51

fascinating, obviously, in an organization like

20:53

yours to be able to have

20:55

that role. of have some influence

20:58

on the product as a peer

21:00

to many of the company's customers.

21:02

Talk about what you see is

21:04

the ongoing evolution of that program.

21:06

Yeah, I mean, you know, I wish I had

21:08

more influence on the product actually. It's,

21:10

you would think that we were first

21:12

in line, we're actually kind of last,

21:14

right? I mean, our paying customers are

21:17

pretty important to us as a company.

21:19

So, you know, part of what we're

21:21

doing is being better customers of our

21:23

product organization. going really

21:25

big with that being really really

21:27

good partners for the folks that

21:29

are building and putting out our

21:32

product. Moreover though it's about going from

21:34

can we do this to yeah I

21:36

think this is going to work for us

21:38

to going like really bombastic with

21:40

it. So you know internally we're

21:42

starting to talk about it instead of

21:44

customer zero we're talking about it as

21:47

like building the operating system. So that's

21:49

what we put out there in our

21:51

marketing materials, what we tell our customers

21:53

to do, right? You can run your

21:55

entire enterprise on Palantier Foundry. And so

21:57

we're like, okay, let's see what that looks like in.

22:00

Let's see how far we can push that envelope

22:02

in 25 and 26. So, you know, 23 for

22:04

us was about pivoting away from the

22:06

traditional tech stack and starting to change

22:08

people's training and how they're approaching the

22:10

work and making it part of everybody's

22:13

job to use our products in a

22:15

different way than they had before. 24

22:17

was about going to production with that,

22:20

getting that flywheel spinning faster, getting

22:22

a few really big Marquis use cases

22:24

out there, starting to really generate

22:26

some. some energy and some attention

22:28

with our customers and internally. 25

22:30

is about like this is a

22:33

mature operating model. This is what

22:35

we're going to be doing. We're

22:37

going really big. And that's fascinating.

22:39

Talk of it about the ongoing

22:41

reshaping of the team to a

22:44

greater extent. Sure. That's what will

22:46

that's kind of the other part

22:48

of of building the operating system.

22:50

So the traditional staffing model of

22:53

the traditional like Well, go

22:55

back to 2014, 15, 16,

22:57

right? At that time, Palantir's

22:59

headcount was accelerating really rapidly,

23:01

many other tech companies was

23:03

like, you know, great, like, headcount growth,

23:05

you were all us, right, to unlock

23:08

growth and you're in your revenue and

23:10

whatever. So at the time we went,

23:12

we went, full sass, full cloud, like

23:14

everybody else did, right? And so there's

23:16

a bunch of assumptions that come along

23:19

with that about what the team ought

23:21

to look like. Going to the you

23:23

have a bunch of specialized skills, right?

23:25

You have sales force engineers, you have

23:27

work day engineers, you have net suite

23:30

engineers, these people like, they're specialists in

23:32

these platforms, it's largely the same type

23:34

of programming language that they do, but

23:36

like with proprietary little tweaks on it,

23:39

they carry industry certifications for these things

23:41

and they're functional and techno functional and

23:43

like there's like a specific array of

23:46

like the way that you build these teams.

23:48

You supplement them with. contractors or

23:50

consultants from a specific set of houses

23:52

that do a specific set of things.

23:54

And there's just kind of like a,

23:56

there's like this diriger kind of like de

23:58

facto way that you, you staff. this and run

24:01

it. As we adopt more foundry

24:03

as the operating system for IT,

24:05

we're realizing that there's a generalizable

24:07

skill set that's happening here.

24:10

So that we have really skilled

24:12

systems engineers from all of these

24:14

different SAS worlds. You're still just

24:17

strong mainstay members of the team

24:19

that this expertise is there for

24:21

reason. As they learn to. address

24:24

more of that engineering thinking to

24:26

the foundry platform and the AI

24:29

assist and co-pilot features in the

24:31

platform, help them do that and

24:33

accelerate their learning and output.

24:36

We're finding that they can, that

24:38

there's a generalizable thing where you

24:40

really only need one or

24:42

two subject matter experts surrounded

24:45

by a bunch of found

24:47

like really good foundry systems

24:49

engineers. And so it's letting us be

24:52

a lot more of the way our

24:54

business site is where like there's famously

24:56

for a long time there were like 2,000

24:58

people with the title for deployed engineer

25:00

and Palantir, right? And it was so

25:02

that everybody could flow and flex around

25:04

and like go to wherever the need

25:07

was or whatever. What we need to

25:09

be able to do in IT is

25:11

that. And so that's the thing that

25:13

we're trying to finish the stroke on

25:15

in 25. Take those generalizable skill sets,

25:18

build these flexible fluid teams so that

25:20

way we can all jump on the

25:22

most important outcomes, surge on them, and

25:24

bang them out that much better. That makes

25:26

sense. And you also are hoping to

25:28

harden the systems and foster greater

25:30

degrees of resiliency makes a lot

25:32

of sense, but what are some

25:34

of the methods you're doing to

25:36

ensure that's the case? Yeah, yeah,

25:38

this is this pretty different topic

25:40

than we've talked about so far in

25:42

the conversation, but you know,

25:44

every company needs to be

25:47

thinking about resilience here, right

25:49

now, if you look at the way the world

25:51

is, it's dynamic, right? And at

25:54

the very least, as dynamic and

25:56

uncertain as things are, that behooves

25:58

you to think about. how do

26:00

you keep your company running? For

26:02

many CIOs also, I would argue,

26:05

we should be thinking about how

26:07

to position our companies for maximum

26:09

advantage in a dynamic

26:11

situation. And certainly in Palantir,

26:13

that's the case. Even in calm

26:15

times, part of my job is to design

26:17

and plan for maximum optionality

26:20

for our business decision makers,

26:22

not for tiny margin optimization,

26:24

right? Like we really think

26:27

about. keeping the aperture of decision-making

26:29

open as far into the future

26:31

as possible and not letting tech

26:33

choices or like the purity of

26:35

a roadmap lock us into old

26:37

assumptions. So in a dynamic world

26:39

like we have in 25, there's just an

26:41

embarrassment of riches of things

26:43

that the CIO ought to be thinking about

26:45

or that I'm thinking about, right? So it's

26:48

like, just to name a few in no

26:50

particular order. Industrial sabotage is

26:52

a really huge one, right? The

26:54

next time somebody drags their anchor

26:56

over undersea fiber somewhere in the

26:59

world. The specter of trade imbalances

27:01

or tariffs or difficulty getting things,

27:03

whatever you want to call it

27:06

today, and by the time this

27:08

comes out, I'm sure the news

27:10

will be totally different than it

27:12

was today. Yesterday was different than

27:14

that. Certainly the... ongoing escalation

27:16

of cyber warfare, information security

27:19

between the United States and

27:21

allied countries and some others

27:23

is getting more and more aggressive,

27:26

therefore moving people around the world,

27:28

is getting and keeping them connected,

27:30

keeping them safe, is much more challenging.

27:33

And like a lot of companies, you

27:35

know, we're all AWS shops or maybe

27:37

we were Azure shops or whatever, and

27:39

there's always more that infrastructure

27:41

thinkers can be doing. to shore up

27:44

like what's your what's your multi-region?

27:46

What's your multi-a-z zone? Multi-a-z strategy?

27:48

I think there's a lot of

27:51

thoughts out there, but I mean, if

27:53

you like on Gardner's CIO tool,

27:55

it's always in the perennially top

27:57

three things that people are searching.

28:00

I think there's a good reason why.

28:02

Makes sense, certainly. It sounds like in

28:04

the balance, the things you're planning for

28:06

also, or ensuring the organization, your organization,

28:09

is supporting this remarkable growth that the

28:11

company foresees, that certainly has experienced and

28:13

foresees, such that it can do so

28:16

with great nimbleness and ensuring that. You

28:18

know, you can pivot quickly as opportunity

28:20

presents itself, pivot away from danger more

28:23

rapidly as well, and ensure that yours

28:25

is an organization that's fostering that nimbleness.

28:27

Is that fair? Exactly. That's really well

28:30

said. Honestly, that is a good summation

28:32

of, that's one facet of a summation

28:34

of basically everything we've talked about in

28:37

this conversation, right? Like everything that we

28:39

do here has to be with an

28:41

eye toward that nimbleness and that the

28:43

ability to pivot toward a good business

28:46

outcome, pivot toward or away from a

28:48

geopolitical issue, usually toward, we tend to

28:50

run toward them, not away from, at

28:53

Palantir, like that has to be infused

28:55

in everything that we're doing. Absolutely. That

28:57

makes sense. Well, I wanted to, as

29:00

I mentioned in the introduction, you have

29:02

a really unusual background, just hearing you

29:04

speak and the technical details that are

29:07

You know, obviously clearly in your mind

29:09

in an organization that is extraordinarily complex,

29:11

other than the fact that I mentioned

29:14

it before this conversation began, one would

29:16

certainly be very surprised to though that

29:18

when you were at the age of

29:21

30, you were not a technologist at

29:23

all. In fact, to that point, you

29:25

would pursue music, you have a graduate

29:27

degree in it, broke from a university

29:30

that happens to also have a great

29:32

technology program in addition to a great

29:34

music program in Carnegie Mellon, But at

29:37

30 you decided you might try something

29:39

else and talk a bit. Talk a

29:41

bit about that pivot from music to

29:44

what would eventually be the latter that

29:46

you began to climb at Palantir. Yeah.

29:48

Yeah, there's so many directions to be

29:51

this. I joke that a lot of

29:53

the reason why I ever got any

29:55

job in technology is because I accidentally

29:58

got music degrees from great engineering schools,

30:00

not with that as the goal. But

30:02

because I was in those school, I

30:04

went to the University of Illinois for

30:07

my undergrad and Carnegie Mellon for a

30:09

master's degree. And because I was just,

30:11

yeah, I was doing this thing, I

30:14

was hanging out with a bunch of

30:16

engineers for musicians for a hobby and

30:18

I was like a. a musician as

30:21

a profession who sort of like dabbled

30:23

in being an engineer for a hobby,

30:25

right? Like I was always sort of

30:28

curious about this stuff. It was always

30:30

exposed to world-class engineers, engineers and computer

30:32

scientists and roboticists and so forth. And

30:35

so the music business, especially in the

30:37

early 2000s with the Great Recession at

30:39

that time and other sort of economic

30:42

pickups, it's a tough business. It's always

30:44

a tough business, right? And for every

30:46

one person used in the Boston Symphony,

30:48

there are hundreds of the rest of

30:51

us who are teaching high school lessons

30:53

and spending a lot of time in

30:55

the car driving from place to place.

30:58

And I got to my late 20s

31:00

and thought, I don't really like this

31:02

a whole lot anymore. And it's one

31:05

of those businesses. You really got to

31:07

be in it. And I have all

31:09

these other curiosities and I've always had

31:12

all these other curiosities. Let me think

31:14

about doing something different, right? And somebody

31:16

who's really hungry for these seats in

31:19

the Youngstown Symphony in Ohio or whatever,

31:21

that I'm just taking up a seat

31:23

not really liking it anymore, somebody else

31:25

who's really hungry for it can have

31:28

that gig. So I was working at

31:30

an Apple store at the time, at

31:32

the Genius Bar and Apple store, by

31:35

the way, our head of IT operations

31:37

now. was my lead at that store

31:39

at that time. We stayed in touch

31:42

all these years and I ended up

31:44

finally stealing him to come to Palantier.

31:46

So, and I had a customer there

31:49

who I was working with who had

31:51

a Palantier like a recruiting. swag water

31:53

bottle and I had never heard of

31:56

Palants here in those days this is

31:58

2012 outside of the you know intelligence

32:00

community or you know some some small

32:02

set of financial services nobody really had

32:05

right it wasn't it wasn't a big

32:07

name wasn't you know the notoriety wasn't

32:09

there yet so company sounded interesting I

32:12

went online and I sort of to

32:14

the jobs I wasn't applying for any

32:16

other jobs I had a perfectly fine

32:19

job at this Apple store And so

32:21

I picked the only thing that that

32:23

I could apply for that I thought

32:26

I wasn't going to laugh out of

32:28

the room for which was tech support

32:30

in the only place that there was

32:33

an opening at that time, which was

32:35

the Washington DC office. I was living

32:37

in Pittsburgh. So it's easy. It's four

32:40

hours down the road and applied and

32:42

didn't get the job. I was offered

32:44

a contractor. So that was that was

32:46

February of 2013. Do you do you

32:49

do you are there? Aspects of your

32:51

time as a musician and the way

32:53

in which that's oriented your mind that

32:56

you find useful in work that you

32:58

do. That is a fantastic question. Yes,

33:00

every day. Classical musicians don't say this

33:03

out loud very often, but it's really

33:05

like a craftsmanship job, more than an

33:07

artistry job to sit in the back

33:10

of an orchestra. So, you know, it's

33:12

like you're, nobody showed up, I was

33:14

a trombonist, right? Nobody pays. money for

33:17

a ticket to hear the guy in

33:19

the trombone section moat. They came to

33:21

hear what what Wagner or Beethoven or

33:23

Brahms wrote, right? So there's a piece

33:26

of spending hours and hours every day

33:28

to align with this other vision. reproduce

33:30

it extremely skillfully and kind of maybe

33:33

put your own spin on it every

33:35

once in a while. That sounds an

33:37

awful lot like the IT business in

33:40

a lot of ways, right? So that's

33:42

one thing. The pain tolerance I think

33:44

equipped me really well for Palantir, where

33:47

you spend eight hours, you know, tens

33:49

of thousands of hours of your life

33:51

for free, just getting yelled at, just

33:54

practicing to make yourself better and torturing

33:56

yourself to get there. That sounds a

33:58

lot like Palantir. in some ways. And

34:01

I think, you know, for me personally,

34:03

once I had been on stage

34:05

in front of, you know, giant

34:07

paying audience and playing a solo

34:09

and that I was unprepared for

34:11

in a weird, you know, these

34:13

weird circumstances that arise

34:15

in the performing arts, I was

34:18

pretty rattle-proof. Like, you know,

34:20

no business executive is

34:22

as scary as name your

34:24

famous orchestra conductor, right? Like, so,

34:26

yeah, I think it was actually...

34:28

a great preparation and there are

34:30

a lot of other Palestinians who have

34:33

come from the liberal arts not

34:35

necessarily music although there are a few

34:37

most of the musicians stay in music

34:39

right like they didn't flunk out like I

34:41

did but there are a lot of people

34:44

from the liberal arts there are a lot

34:46

of people from the humanities because I think

34:48

there's a there's a lot of them have

34:50

that in common right there's that like

34:52

perfecting your craft there's this thinking a

34:55

very specific way that like leads you

34:57

to to communicate and collaborate in a

34:59

specific way. And a lot of those

35:01

folks, especially in IT in my world,

35:03

I prefer folks like that, that come

35:06

from these unconventional backgrounds. They

35:08

bring a lot of interesting thinking.

35:10

They're super interesting. Thank

35:12

you for those reflections. I wanted

35:14

to ask, we've talked about a number of

35:17

topics throughout this, any trends as you

35:19

look to the future, the year ahead

35:21

or even beyond, or have you particularly

35:23

intrigued, what would you, what would you

35:25

underscore Jim? I mean the elephant in

35:27

the corner of their room is AI, right?

35:30

Like every, you know, every LinkedIn thought leadership

35:32

post and every, you know, whatever, it's

35:34

all about what is AI mean? Is it

35:36

worth anything? Is it not worth anything? Is

35:38

it not worth anything? Is it not worth

35:41

anything? Is agentic architecture going to be that,

35:43

you know, so I think that the growing

35:45

trend that I'm excited about is the specific

35:48

viewpoint, quiet, quiet and small at

35:50

first and getting louder and louder

35:52

as it produces the best results

35:54

that You cannot get

35:57

and go by an

35:59

AI thing. because you're bored yelled at

36:01

you to not miss the bus, and then

36:03

run around with a hammer looking for

36:05

nails trying to find value with it.

36:07

There's an increasingly rigorous thinking about how

36:10

to apply AI in a way that's

36:12

valuable for IT because you're

36:14

already running an IT organization that's

36:16

valuable for your business. And

36:18

so I think it's forcing a lot of

36:21

us in the IT profession to really

36:23

think critically about what it is we're

36:25

actually trying to do. to run our

36:27

businesses, not just to support, but to

36:29

actively be accretive to the business and

36:32

actually value building and be worth an investment.

36:34

It's a knock on benefit to AI that

36:36

I think a lot of people, you talked

36:38

about, like labor savings and all this

36:40

other stuff, but I think it's increasing

36:42

focus on the value of IT, I

36:44

think, is something that we're under talking

36:46

about. I think it's just growing. That's

36:48

great. Great points you raise. I wanted

36:50

to also ask you, is there anything

36:53

you've recently read, watched, and or listened

36:55

to that you would recommend to those

36:57

who are watching and listening to this?

37:00

Yes, I'll give you one of all

37:02

three. On watching, I've just got done

37:04

re-watching Star Trek The Next Generation.

37:07

It's very serious scholarly work.

37:09

I find myself quoting Captain Picard

37:11

far too often, poorly, poorly quoting.

37:14

I've just gotten done rereading after

37:16

I tend to reread and rewatch

37:19

things rereading after many years probably

37:21

10 years since I last reread

37:23

it. I'm a history buff. I just

37:25

I got done reading by Doris

37:27

Kearns Goodwin the book Team of

37:29

Rivals about Abraham Lincoln and how

37:31

Abraham Lincoln built his cabinet and

37:34

the really complex dynamics that they

37:36

had. Every time it is my

37:38

third time reading this book over

37:40

the course of probably 15 years,

37:42

10 15 years and it's and it's

37:44

Every time I read it I learn

37:46

something else. It's why I keep rereading

37:48

things that are worth rereading. I just,

37:50

it's amazing. And as far as

37:52

podcasts go, I recently got really

37:55

into the podcast acquired. Just, I

37:57

think I was late to that party. I

37:59

think every... else in the business was already

38:01

into it and it was just like a

38:03

bombshell to me it's like this long-form

38:06

sort of the adventure story of

38:08

running a business like fantastic I just

38:10

I find it amazing like an audio

38:12

book in conversation between the two yes

38:14

absolutely absolutely great great suggestions all around

38:17

Jim and Jim thanks it more over

38:19

for a great conversation across the board

38:21

really fascinating here more about your career

38:23

journey more about the evolution of the

38:26

company and the the the many steps

38:28

that you're taking to ensure that yours

38:30

is a team that's supporting the remarkable

38:33

growth that this company has had and

38:35

likely will continue to have as well.

38:37

It's been a great conversation. Thank you,

38:40

Peter. Thank you for having me. It's been

38:42

a great time.