Transcript
1
00:00:00,017 --> 00:00:05,017
Welcome back. Sandy here, and I have some good news and some bad news to share about this episode.
2
00:00:05,257 --> 00:00:09,817
The good news is we have a guest joining us. The bad news is he's a vendor.
3
00:00:09,977 --> 00:00:14,137
But all jokes aside, Anjali and I had a fantastic time chatting with James Anderson.
4
00:00:14,457 --> 00:00:18,297
He's an ex-consultant and current director of cloud data architecture at Snowflake.
5
00:00:18,437 --> 00:00:22,657
James and I go way back, and I've always enjoyed our spirited debates,
6
00:00:22,937 --> 00:00:27,097
which is exactly why Anjali and I decided to invite him to the podcast. guests.
7
00:00:27,137 --> 00:00:29,937
We couldn't think of a livelier guest to kick things off with.
8
00:00:30,077 --> 00:00:34,457
Yes, we start by discussing Snowflake, of course, but then we move into other
9
00:00:34,457 --> 00:00:38,877
topics like data mesh, changing dynamics of dashboard usage and data analytics.
10
00:00:39,137 --> 00:00:43,477
And we have a quick debate on the future of data work. We hope you enjoy listening
11
00:00:43,477 --> 00:00:46,417
to this discussion as much as we enjoyed having it. So let's get started.
12
00:00:46,320 --> 00:01:05,840
Music.
13
00:01:06,163 --> 00:01:11,163
Nice. You know when your recording's on and you have the little red flashing
14
00:01:11,163 --> 00:01:13,403
light back in the studio back in the day? Uh-huh.
15
00:01:14,443 --> 00:01:17,103
This is outrageous. And I like that
16
00:01:17,103 --> 00:01:19,843
we're recording now, this part of this conversation. It makes me happy.
17
00:01:21,343 --> 00:01:24,363
I hope this makes it into the podcast. You said I didn't hit the button earlier. That's what I'm...
18
00:01:26,783 --> 00:01:29,703
All right. Are you mentally prepared? You want to get into this?
19
00:01:30,143 --> 00:01:32,703
Am I mentally prepared? I was on a red-eye flight last night.
20
00:01:32,783 --> 00:01:35,303
I'm never mentally prepared. So let's just... Let's roll.
21
00:01:35,683 --> 00:01:40,743
Red Eye. Where did you come from? San Francisco. I've been flying around a lot. Back and forth to HQ.
22
00:01:40,983 --> 00:01:44,243
So James, I've known you for a while now.
23
00:01:44,543 --> 00:01:49,623
For a long while. I don't even know how long. But how about you? Almost 10 years.
24
00:01:49,843 --> 00:01:52,903
I've been at Cervelo for 11. So that's kind of incredible.
25
00:01:53,103 --> 00:01:58,883
Now that I think about it. I started working at Cervelo in November of 2014.
26
00:01:59,403 --> 00:02:02,243
I didn't realize that for some reason. But why don't you introduce yourself
27
00:02:02,243 --> 00:02:05,183
to Anjali and our listeners while you're at it? Sure.
28
00:02:05,443 --> 00:02:09,943
Yeah. So my name is James Anderson. I used to work with Sandy and the team at
29
00:02:09,943 --> 00:02:12,863
Cervelo for I think it was about two years I was there.
30
00:02:12,983 --> 00:02:15,243
I then moved on to another consulting firm.
31
00:02:15,363 --> 00:02:18,063
You know, I started my career at Cervelo as a burst developer.
32
00:02:18,303 --> 00:02:24,183
I was a front end front end dashboard guy building very beautiful dashboards in burst as one one does.
33
00:02:24,343 --> 00:02:29,583
And then as I moved in through my consulting career, I moved myself down the stack.
34
00:02:29,723 --> 00:02:34,023
And by the time I left the consulting space, I was running the platform architecture
35
00:02:34,023 --> 00:02:38,883
team for the firm I was at focused on how do you build and deploy large scale
36
00:02:38,883 --> 00:02:40,603
enterprise wide data platforms?
37
00:02:40,763 --> 00:02:43,243
And how do you fit that into a broader enterprise architecture,
38
00:02:43,483 --> 00:02:48,043
I left consulting in 2020, right at the height of the pandemic moved over to
39
00:02:48,043 --> 00:02:51,163
snowflake because I had been doing almost exclusively snowflake implementations
40
00:02:51,163 --> 00:02:54,623
for at least the previous four years before that.
41
00:02:54,703 --> 00:02:57,443
So I've been in and around the snowflake ecosystem for a long time.
42
00:02:57,543 --> 00:03:00,903
And I have been at snowflake now for just shy of four years.
43
00:03:01,183 --> 00:03:04,703
For the first couple of years, I ran a number of different sales engineering
44
00:03:04,703 --> 00:03:08,603
teams focused on first large accounts in the greater New England area.
45
00:03:08,803 --> 00:03:12,143
And then for the last two years, I had been running the sales engineering team
46
00:03:12,143 --> 00:03:14,663
focused on large life sciences organizations.
47
00:03:15,083 --> 00:03:19,423
So pharmaceutical, medical device, medical distribution, CRO.
48
00:03:19,643 --> 00:03:24,823
And starting February, I moved into our new data cloud architecture team focused
49
00:03:24,823 --> 00:03:28,743
on how do you fit Snowflake into to a more broad enterprise architecture.
50
00:03:29,023 --> 00:03:33,743
That's my complete end-to-end life that I have had in the data space at this
51
00:03:33,743 --> 00:03:35,663
point. What do you do outside of the data space?
52
00:03:35,883 --> 00:03:40,703
What kind of hobbies or interests do you have? Or do you work so hard you have no hobbies, interests?
53
00:03:40,803 --> 00:03:46,043
Well, in the last month or so, let's see, I've been on 10 planes in the last 20 days.
54
00:03:46,243 --> 00:03:49,943
So I don't really have time for other things outside of that right this second.
55
00:03:50,223 --> 00:03:53,723
In general, I have two young children who I adore.
56
00:03:54,314 --> 00:03:58,314
Immensely and spend most of my time entertaining or finding ways to entertain.
57
00:03:58,514 --> 00:04:02,594
I also like to play golf like any good data executive likes to do.
58
00:04:02,794 --> 00:04:05,874
Those are my main hobbies if I have time for those things.
59
00:04:06,214 --> 00:04:09,754
And we'll see if that ever comes back, if I ever get more time for that.
60
00:04:09,994 --> 00:04:13,134
So is that a newly formed team at Snowflake? Yeah.
61
00:04:13,194 --> 00:04:17,434
So historically, when our customers had come to us and said,
62
00:04:17,514 --> 00:04:20,194
hey, what is our recommended architectural approach?
63
00:04:20,734 --> 00:04:24,754
Hey, should we run a data warehouse? Should we transition to to being more of
64
00:04:24,754 --> 00:04:26,134
a data lake type of construct?
65
00:04:26,954 --> 00:04:31,354
What do you guys think about data fabric? Definitely getting a lot of people asking about data mesh.
66
00:04:31,594 --> 00:04:34,694
In the past, our answer was, oh, you can do anything you want.
67
00:04:34,814 --> 00:04:40,154
As an organization, we're really trying to get our customers to think about
68
00:04:40,154 --> 00:04:44,794
Snowflake, not just as a cloud data warehouse or cloud data lake or something
69
00:04:44,794 --> 00:04:48,494
of that nature, but really understand that this is a full-blown end-to-end data platform.
70
00:04:48,674 --> 00:04:53,554
And so we've built an architectural framework that we call the data cloud architecture,
71
00:04:53,874 --> 00:04:59,254
to help our customers who want to truly adopt Snowflake as a platform and get
72
00:04:59,254 --> 00:05:01,514
better value out of the data that they have.
73
00:05:01,714 --> 00:05:07,614
A lot of this framework is around how do you tie value to what you're building?
74
00:05:07,754 --> 00:05:11,574
And how do you treat the things that you're building not as tools for decision
75
00:05:11,574 --> 00:05:15,654
making, but instead as an asset that has a measurable ROI?
76
00:05:16,094 --> 00:05:21,654
In the past, when a client says, says, hey, where does this fit in my larger architecture?
77
00:05:22,134 --> 00:05:29,174
The answer would naturally be go to consultant X who's worked across a lot of technologies, right?
78
00:05:29,354 --> 00:05:33,574
To help you pull that together. So now it seems like Snowflake's building a
79
00:05:33,574 --> 00:05:36,054
team that has an opinion, which is fine.
80
00:05:36,214 --> 00:05:40,594
Find it interesting to see how clients react to that because you're still getting
81
00:05:40,594 --> 00:05:42,774
the advice from the vendor at the end of the day.
82
00:05:42,854 --> 00:05:46,494
Yeah, so one of the things that I tell customers from the jump.
83
00:05:46,842 --> 00:05:52,662
Is that my team is certainly going to look at this from a snowflake first perspective,
84
00:05:52,902 --> 00:05:55,862
but we're not going to look at it from a snowflake only perspective.
85
00:05:56,022 --> 00:06:01,602
We are a little bit trying to provide some level of free consulting to help lay that foundation.
86
00:06:01,782 --> 00:06:04,602
Now, we're not going to then turn around and do your whole implementation for
87
00:06:04,602 --> 00:06:06,562
free for you for all the way through, right?
88
00:06:06,642 --> 00:06:11,682
We're going to rely on our partners like Cervelo to actually take this sort
89
00:06:11,682 --> 00:06:14,362
of architectural vision and implement it and put it in place.
90
00:06:14,442 --> 00:06:17,962
Because there's a lot, It's not just a technical underpinnings, right?
91
00:06:18,022 --> 00:06:22,282
There's a shift in how you think, and there's a change management process that
92
00:06:22,282 --> 00:06:23,882
has to be taken into account.
93
00:06:24,182 --> 00:06:29,342
So as you're talking about this, I guess because Snowflake made this shift,
94
00:06:29,662 --> 00:06:32,162
it sounds like that's been a big trend, right?
95
00:06:32,262 --> 00:06:37,682
Clients really trying to figure out not just what is this technology itself
96
00:06:37,682 --> 00:06:43,122
in isolation, but how does it interplay with all the other things I'm trying to do in my stack?
97
00:06:43,122 --> 00:06:47,582
The question you're trying to answer for them is, here are the other pieces
98
00:06:47,582 --> 00:06:49,662
of my stack. Here's the other things I'm trying to achieve.
99
00:06:50,002 --> 00:06:56,902
Help me put this wonderful modular puzzle together that has been created as
100
00:06:56,902 --> 00:06:59,882
the modern, quote unquote, modern stack.
101
00:07:00,002 --> 00:07:02,602
It's modular, right? That's usually what people push for.
102
00:07:02,882 --> 00:07:08,722
I know companies are moving in certain directions, but that's typically what people push for today.
103
00:07:09,022 --> 00:07:11,962
Are there any other trends that you're seeing out there? I mean,
104
00:07:12,042 --> 00:07:15,762
data mesh is the most popular conversation that we've had.
105
00:07:15,882 --> 00:07:19,022
I mean, is it not the most popular conversation you've had in the last year
106
00:07:19,022 --> 00:07:20,262
and a half? Like, really? No.
107
00:07:20,782 --> 00:07:25,962
Really? Yeah. That's surprising, because the number of customers that have come
108
00:07:25,962 --> 00:07:31,042
to me and said, Hey, I have this partner who's really pushing the data mesh paradigm.
109
00:07:31,462 --> 00:07:35,142
What do you guys think? And how can Snowflake support this data mesh strategy?
110
00:07:35,402 --> 00:07:40,702
Right? Is outrageous. I think last summer I had this conversation with more
111
00:07:40,702 --> 00:07:42,282
than half of the customers that I supported.
112
00:07:42,382 --> 00:07:48,882
Everybody likes this idea of domain ownership and putting the capabilities and
113
00:07:48,882 --> 00:07:52,162
the infrastructure in the hands of the people who actually understand the data.
114
00:07:52,162 --> 00:07:56,262
On the flip side, when it comes to a true data mesh, the idea of domain ownership
115
00:07:56,262 --> 00:08:01,782
includes the idea that you have engineers inside of each domain who will take
116
00:08:01,782 --> 00:08:05,662
who will ingest process and create these data products.
117
00:08:05,862 --> 00:08:09,422
And I have a whole beef on the term data product, which we can get into later.
118
00:08:09,582 --> 00:08:14,562
But there's a level of sort of organizational change that goes into how you
119
00:08:14,562 --> 00:08:17,242
staff. Yeah, or how you organize period, right.
120
00:08:17,322 --> 00:08:20,262
So from our frameworks perspective, from this sort of data cloud architecture
121
00:08:20,262 --> 00:08:23,842
perspective, we're taking a lot of the positives of data mesh in terms of,
122
00:08:23,842 --> 00:08:27,502
we want to enable that free flow of data and applications across your ecosystem,
123
00:08:27,682 --> 00:08:32,162
we want to make it easier for you to collaborate with your business and with your partners.
124
00:08:32,763 --> 00:08:36,423
And with your customers even, but not do it in a way where you have to actually
125
00:08:36,423 --> 00:08:37,803
transform how you run your business.
126
00:08:37,943 --> 00:08:40,583
And I understand that there are consulting firms out there who specialize in
127
00:08:40,583 --> 00:08:41,243
business transformation.
128
00:08:41,503 --> 00:08:45,423
So this is very appealing to them, because there's a huge project that comes behind this.
129
00:08:45,783 --> 00:08:49,423
So I think that's one. The other is obviously generative AI and everything around
130
00:08:49,423 --> 00:08:50,783
the whole generative AI space.
131
00:08:51,063 --> 00:08:54,503
And I know when I was listening to your first episode, I was texting Sandy with
132
00:08:54,503 --> 00:08:56,343
all sorts of opinions on this.
133
00:08:56,483 --> 00:08:59,783
Well, that's how you ended up on the podcast. You were texting me right after
134
00:08:59,783 --> 00:09:03,983
our first episode, So you're sending me these rant texts about Gen AI.
135
00:09:04,203 --> 00:09:07,083
And I was like, wow, we got to get him on. He's ready to go.
136
00:09:07,543 --> 00:09:11,723
I was live texting you as I was listening with my live feedback. Yes.
137
00:09:12,463 --> 00:09:15,163
But would you expect me to do it any other way, Sandy? I mean,
138
00:09:15,183 --> 00:09:17,283
come on. No, absolutely not. That's what I love about you, James.
139
00:09:17,383 --> 00:09:21,943
I think when it comes to data mesh, yes, we do have clients who are thinking about it.
140
00:09:22,003 --> 00:09:25,103
We don't push it as a consulting firm. And I think because we know,
141
00:09:25,203 --> 00:09:29,763
we know what it takes. And when I can't get a company to even think about how
142
00:09:29,763 --> 00:09:34,643
to model data for consumption, why am I going to start pushing a concept that they're not ready for?
143
00:09:34,743 --> 00:09:37,043
As a consultant, I'm not even going to talk about our company,
144
00:09:37,163 --> 00:09:40,163
but as a consultant, I personally have a challenge with that.
145
00:09:40,443 --> 00:09:45,963
With the push of new concepts to clients when they are not mature enough to
146
00:09:45,963 --> 00:09:49,083
get to that space, that's something that I holistically will not do.
147
00:09:49,403 --> 00:09:53,063
I want people to be successful in what they're trying to achieve.
148
00:09:53,223 --> 00:09:56,403
That's my goal. So yeah, I don't push it personally.
149
00:09:56,583 --> 00:10:00,683
I think data products, I know you don't like that term. I don't like many terms
150
00:10:00,683 --> 00:10:03,583
because people define things however they want to define them.
151
00:10:03,943 --> 00:10:07,723
So I always say, you have your definition, I have mine. Let's just make sure
152
00:10:07,723 --> 00:10:10,283
we all understand where we're coming from is kind of my goal in life.
153
00:10:10,703 --> 00:10:14,643
But data mesh, yeah, we have clients have talked about it.
154
00:10:14,723 --> 00:10:18,763
And most of the time, they're looking at it as a tooling question.
155
00:10:19,183 --> 00:10:23,683
What are the technical pieces I need to create a quote unquote data mesh? And I tell them,
156
00:10:24,085 --> 00:10:28,085
That's nice. That's not going to solve your problem. Well, yeah,
157
00:10:28,265 --> 00:10:32,305
I mean, I agree with the maturity aspect, but there are companies out there
158
00:10:32,305 --> 00:10:33,805
who are structured in this way.
159
00:10:34,065 --> 00:10:37,925
Extremely large e-commerce retailer here in the Boston area is structured that
160
00:10:37,925 --> 00:10:41,285
way, right? That's how their business is set up.
161
00:10:41,385 --> 00:10:43,665
So fine, you know what? That makes sense.
162
00:10:43,865 --> 00:10:47,185
Good for you. Good job. For everybody else, yikes.
163
00:10:48,685 --> 00:10:52,465
So my BeatBoot data product, in my opinion, a data product is something that
164
00:10:52,465 --> 00:10:54,545
you build in order to sell to the rest of your mesh.
165
00:10:54,725 --> 00:10:58,185
You don't know that anybody's going to use it. You just think that it is.
166
00:10:58,225 --> 00:11:02,305
But you have to like spend cycles building it in such a way that it feels like
167
00:11:02,305 --> 00:11:06,645
an actual product, which in my mind actually lowers the ROI that you might get
168
00:11:06,645 --> 00:11:09,585
out of that because your input costs are higher.
169
00:11:09,705 --> 00:11:13,285
Your resource costs are higher. The technology costs that you're using in order
170
00:11:13,285 --> 00:11:15,345
to make this thing look like a product are probably higher.
171
00:11:15,465 --> 00:11:18,785
And you could probably do something with lower input costs and get the same
172
00:11:18,785 --> 00:11:20,065
level of adoption across the board.
173
00:11:20,165 --> 00:11:24,865
That's why we're really focused on this idea of assets sets and tying everything back to ROI.
174
00:11:25,085 --> 00:11:28,025
Because then it's really about are you building the right things?
175
00:11:28,285 --> 00:11:32,065
Like previous consulting firm, we did a project for a large hospital network
176
00:11:32,065 --> 00:11:33,905
in general Massachusetts area.
177
00:11:34,225 --> 00:11:37,965
And we built them a dashboard. It was there was a lot of data that came into
178
00:11:37,965 --> 00:11:40,825
this and it was a pretty hefty project and program that we did.
179
00:11:41,065 --> 00:11:45,025
It was a bed management dashboard where they would be able to see sort of the
180
00:11:45,025 --> 00:11:48,485
layout of how many beds were available and so on and so forth,
181
00:11:48,605 --> 00:11:53,325
and some level of a sort of a light forecast against sort of where the beds
182
00:11:53,325 --> 00:11:56,845
would be, so that they could do their staffing in such a way that they saved
183
00:11:56,845 --> 00:12:01,405
something, they told us they saved something in the realm of like $70 million in staffing costs.
184
00:12:01,525 --> 00:12:05,625
And the fact that they like made an effort to track that number tells me that
185
00:12:05,625 --> 00:12:08,865
they actually understood the ROI that they were going for.
186
00:12:09,025 --> 00:12:12,805
Yeah, I think I look at it similar, but different than the approach you just
187
00:12:12,805 --> 00:12:16,445
laid out, right? I guess, James, like the example you gave, right? It was a.
188
00:12:16,919 --> 00:12:21,799
There was a specific ROI that we were targeting, a very pointed product that was being created.
189
00:12:21,999 --> 00:12:25,739
I look at something like that. I'm like, all right, that's the goal of the business.
190
00:12:25,779 --> 00:12:27,359
That's your business strategy, right?
191
00:12:27,619 --> 00:12:29,639
And ultimately, there's a product that's tied to that.
192
00:12:29,919 --> 00:12:32,999
But I look at all the data elements that are tied to that product.
193
00:12:33,079 --> 00:12:36,599
And I think to myself, this concept of, because this is what happens in organizations,
194
00:12:36,879 --> 00:12:39,559
right? They have this ROI item that they're trying to achieve.
195
00:12:39,879 --> 00:12:42,599
And they're like, all right, we need to get all this work done for this one
196
00:12:42,599 --> 00:12:47,919
use case. And then all the work that is done for that use case is only for that use case.
197
00:12:47,999 --> 00:12:51,659
Everything around it gets orphaned. That big data product probably includes
198
00:12:51,659 --> 00:12:56,559
20 others underneath it, pieces and components of that product that are very
199
00:12:56,559 --> 00:12:59,859
specific to domains that other people can re-leverage.
200
00:12:59,859 --> 00:13:04,539
So you need to distill down what you're actually trying to do into components
201
00:13:04,539 --> 00:13:06,959
so that people can re leverage that information.
202
00:13:07,119 --> 00:13:11,539
And then the ROI goes way up. And not only for that one project that you just
203
00:13:11,539 --> 00:13:15,819
finished on the bedside, but the reuse of those analytics and those different
204
00:13:15,819 --> 00:13:17,619
data products that made that larger one, right?
205
00:13:17,719 --> 00:13:21,039
That's the way I look at it. But the problem is people don't think that way. Right.
206
00:13:21,199 --> 00:13:25,019
Agreed. And this is why, this is why we're, again, we're focused on assets and
207
00:13:25,019 --> 00:13:29,139
why we're focused on ROI, because to your point, there's 20 different other
208
00:13:29,139 --> 00:13:32,739
things that are part of that, that are part that go into building this,
209
00:13:32,739 --> 00:13:34,759
this more broad, larger asset.
210
00:13:34,859 --> 00:13:38,419
This is, and my issue with data mesh becomes that you end up actually building
211
00:13:38,419 --> 00:13:40,959
in silos and then trying to share at the end. Yeah. Right.
212
00:13:41,279 --> 00:13:45,619
Versus like building in a collaborative fashion is is ultimately how you're
213
00:13:45,619 --> 00:13:47,359
going to drive the most efficiencies across the board.
214
00:13:47,479 --> 00:13:51,939
And so part of our goal as a data cloud architecture team is to help to drive
215
00:13:51,939 --> 00:13:56,659
those efficiencies and help reduce that and drive that collaboration by making
216
00:13:56,659 --> 00:14:00,199
it clear who owns what and what's out there.
217
00:14:00,279 --> 00:14:04,399
Anjali, from a data mesh perspective, have you been hearing anything around
218
00:14:04,399 --> 00:14:08,899
that? I hear a little bit of chatter, but I think it's similar.
219
00:14:09,239 --> 00:14:13,799
Sandy, to the approach that you laid out in terms of.
220
00:14:14,516 --> 00:14:16,396
Our clients aren't ready for
221
00:14:16,396 --> 00:14:21,396
kind of what it takes to be successful with data mesh type of approach.
222
00:14:21,676 --> 00:14:27,356
And one of the key foundations for being successful with the data mesh is having
223
00:14:27,356 --> 00:14:32,836
a highly governed approach to your data, setting up a standard to say,
224
00:14:32,936 --> 00:14:36,596
in order to be considered a data product in our mesh,
225
00:14:36,796 --> 00:14:41,916
you need to meet these handful of criteria and actually track against that.
226
00:14:41,916 --> 00:14:45,876
In the consulting space, I always remember governance being the third rail of a proposal.
227
00:14:46,396 --> 00:14:49,636
Don't put governance in there because nobody's going to buy the project or it's
228
00:14:49,636 --> 00:14:51,176
going to be the first thing that gets pulled out.
229
00:14:51,236 --> 00:14:55,056
MDM and governance were the two things that always got stripped out.
230
00:14:55,176 --> 00:14:57,996
It's just like, I just don't even touch it. Those who toil in the governance
231
00:14:57,996 --> 00:15:01,396
space on the consulting side, I feel the most sympathy for ever.
232
00:15:01,556 --> 00:15:05,536
It's not- You're talking directly to Anjali. I know. That's why I'm talking.
233
00:15:05,796 --> 00:15:10,536
I'm sending this at Anjali and telling her how much, how bad I feel for her
234
00:15:10,536 --> 00:15:12,196
in this particular scenario.
235
00:15:13,096 --> 00:15:19,376
Thank you for that. You know, a lot of our clients are actually struggling with,
236
00:15:19,416 --> 00:15:24,356
you know, with data challenges and have opened up their eyes to the need for
237
00:15:24,356 --> 00:15:30,036
finding accountability and focusing on the confidence and fidelity of their data.
238
00:15:30,096 --> 00:15:34,956
So I think that as we kind of look forward, maybe governance will no longer
239
00:15:34,956 --> 00:15:39,576
be that forgotten forgotten child and really be the golden child of the family.
240
00:15:39,876 --> 00:15:42,816
You know, back to the other topic that's been very popular around generative
241
00:15:42,816 --> 00:15:44,796
AI and everybody needing an AI strategy.
242
00:15:45,016 --> 00:15:50,796
So at our summit last year, Frank Slootman, our former CEO on stage said, you can't have.
243
00:15:51,857 --> 00:15:55,257
An AI strategy without a good data strategy. It doesn't surprise me,
244
00:15:55,357 --> 00:16:00,697
Anjali, that governance is less of a problem and people are looking at it more
245
00:16:00,697 --> 00:16:02,957
because they're being asked to build an AI strategy.
246
00:16:03,217 --> 00:16:08,477
And you can have the best LLMs and the best generative AI capabilities and the
247
00:16:08,477 --> 00:16:09,597
best chat bots in the world.
248
00:16:09,677 --> 00:16:13,137
But if your data quality isn't there, you're training models against trash.
249
00:16:13,357 --> 00:16:15,037
So you're going to get trash back.
250
00:16:15,377 --> 00:16:18,377
The good thing about Gen AI is that it's
251
00:16:18,377 --> 00:16:21,477
highlighted highlighted the fact that good quality sound
252
00:16:21,477 --> 00:16:24,737
data and mastering of metadata is
253
00:16:24,737 --> 00:16:27,537
critical to be able to do something like that right
254
00:16:27,537 --> 00:16:30,597
now everybody wants it because it looks so cool and i'd love
255
00:16:30,597 --> 00:16:33,337
to chat with my data someday hopefully we'll get there
256
00:16:33,337 --> 00:16:36,857
you can't get there if your data is out of quality you can't get there if your
257
00:16:36,857 --> 00:16:41,977
your context around your data is aloof right this is the conversation i had
258
00:16:41,977 --> 00:16:47,097
yesterday with the ceo everybody wants to everybody wants a chatbot And I refuse
259
00:16:47,097 --> 00:16:50,217
because they don't trust the data underneath it. And then they all go off and
260
00:16:50,217 --> 00:16:51,297
build in their own little world.
261
00:16:51,417 --> 00:16:55,877
And, and then there's four different definitions of ACV or whatever it is that
262
00:16:55,877 --> 00:16:59,877
they're looking at and revenue and or, you know, four definitions of churn and none of them are right.
263
00:16:59,957 --> 00:17:03,597
And then when you ask them, well, how do you define that? I would say maybe
264
00:17:03,597 --> 00:17:05,817
not none of them are right. Maybe all of them are right.
265
00:17:05,917 --> 00:17:09,617
It's just a matter of the intent behind it. What's my intent?
266
00:17:09,777 --> 00:17:10,577
What am I trying to drive?
267
00:17:10,777 --> 00:17:15,597
It's intent, it's context, and it's quality of the data asset itself that you're going after.
268
00:17:15,977 --> 00:17:19,817
Ultimately, none of this is going to work if companies don't bring together.
269
00:17:20,337 --> 00:17:25,157
Governance with the data teams and the AI teams and the digital teams into one
270
00:17:25,157 --> 00:17:28,697
package and figure out how do we collaborate to solve these problems.
271
00:17:29,325 --> 00:17:33,145
Because if everybody keeps working in isolation, A, they're only going to have
272
00:17:33,145 --> 00:17:34,145
part of the building block.
273
00:17:34,325 --> 00:17:37,285
No one's going to know how to use it. No one's going to trust it.
274
00:17:37,405 --> 00:17:40,605
And it's just going to be a mess again, except on modern platforms.
275
00:17:41,245 --> 00:17:46,745
Well, you keep saying modern platforms. And this is a conversation I've been having as well,
276
00:17:46,865 --> 00:17:53,725
where people in the sort of mid 2010s started to migrate off of their on-premise
277
00:17:53,725 --> 00:17:58,445
data centers into the cloud with the intent of, oh, we built all these silos on-prem.
278
00:17:58,445 --> 00:18:01,865
We're going to move into the cloud and and the silos are just going to magically disappear.
279
00:18:02,445 --> 00:18:07,405
And they picked up and moved their silos yeah and then they kept being siloed in the cloud so
280
00:18:07,825 --> 00:18:12,605
modern stack doesn't always apply if you're if you're not breaking down the
281
00:18:12,605 --> 00:18:15,965
silos and i think a good and you know back to the whole governance aspect like
282
00:18:15,965 --> 00:18:20,085
a good governance is a double-edged sword here right there is such a thing as
283
00:18:20,085 --> 00:18:24,065
too much governance right you put you make it too hard to participate everybody's
284
00:18:24,065 --> 00:18:26,765
going to go off and build in their own silos and shadow it runs runs rampant.
285
00:18:26,785 --> 00:18:31,305
My whole thing around this modern stack has to be in conjunction with a modern
286
00:18:31,305 --> 00:18:33,765
way of thinking, right? Modern ways of working too.
287
00:18:34,125 --> 00:18:38,405
Yes. A hundred percent. Right. And you can't. Yeah. I mean, you're talking about
288
00:18:38,405 --> 00:18:39,705
governance and I agree with you, right?
289
00:18:39,805 --> 00:18:43,745
I look at this and I know Anjali agrees with me as well is governance should
290
00:18:43,745 --> 00:18:47,345
be focused on outcomes and impact that you're trying to achieve.
291
00:18:47,465 --> 00:18:50,725
At the end of the day, it's an enablement factor, right? What am I trying to
292
00:18:50,725 --> 00:18:55,705
enable and building a process that supports that use case or enablement that
293
00:18:55,705 --> 00:18:59,225
you're trying to go forward with, whether it's mastering of data,
294
00:18:59,505 --> 00:19:02,485
you know, reviewing quality and building a process around that,
295
00:19:02,525 --> 00:19:06,285
or even security and masking policies, etc.
296
00:19:06,545 --> 00:19:10,965
But if you're creating processes that restrain the business from doing things,
297
00:19:11,005 --> 00:19:13,405
then you're doing governance for the sake of governance.
298
00:19:13,485 --> 00:19:16,125
And you're actually, you know, not doing it for the sake of the outcome.
299
00:19:16,285 --> 00:19:19,305
And I think people go one way or the other on that. And it's tough,
300
00:19:19,405 --> 00:19:21,905
because the first conversation people typically have is.
301
00:19:22,303 --> 00:19:24,863
Yeah, what's the organization I need to have for governance?
302
00:19:25,143 --> 00:19:28,703
It's just like, I don't want to answer that question. I actually don't care.
303
00:19:28,863 --> 00:19:32,163
The question I want to answer is what do you need to solve for and put the right
304
00:19:32,163 --> 00:19:33,083
process in place for that?
305
00:19:33,243 --> 00:19:37,683
Yeah, I mean, ultimately, governance needs to be an enabler for innovation.
306
00:19:37,843 --> 00:19:40,043
It cannot be the inhibitor to innovation.
307
00:19:40,363 --> 00:19:43,683
It has to be flexible enough to allow for innovation.
308
00:19:43,943 --> 00:19:47,183
Exactly. One question I had, you know, I've been reading a lot about some of
309
00:19:47,183 --> 00:19:51,123
the bold moves that Snowflake has made around MLOps, LLMs, etc.
310
00:19:51,283 --> 00:19:58,763
You guys have kind of taken a shift into the world of advanced analytics, data science, etc.
311
00:19:58,983 --> 00:20:02,423
So I'm curious, what's the thinking behind that?
312
00:20:02,563 --> 00:20:06,523
And how do you envision that shaping the future of data management?
313
00:20:06,663 --> 00:20:10,203
Because historically, right, and I love this idea, by the way,
314
00:20:10,263 --> 00:20:12,663
because I'm all about bringing these worlds together.
315
00:20:12,763 --> 00:20:16,023
As I just said, these teams have to collaborate, they have to be working on
316
00:20:16,023 --> 00:20:19,623
the same platform, not platform necessarily, but the same data sets,
317
00:20:19,763 --> 00:20:21,323
right, the same data products.
318
00:20:21,963 --> 00:20:27,063
So is it really just a shift in that direction of trying to bring these worlds together?
319
00:20:27,283 --> 00:20:30,683
So I think there's a couple of different answers. And I don't,
320
00:20:30,683 --> 00:20:35,963
I don't promise to speak for our product organization and for our leadership in that way.
321
00:20:36,003 --> 00:20:42,603
But yeah, you know, we've made a lot of strategic acquisitions in the last 18 to 24 months, right?
322
00:20:42,683 --> 00:20:46,523
Companies like Neva, which is where Sridhar, our new CEO came from, right?
323
00:20:46,583 --> 00:20:51,983
And And partnerships as well, where we're trying to, again, be a more robust
324
00:20:51,983 --> 00:20:53,323
platform and bring these capabilities.
325
00:20:53,663 --> 00:20:57,563
But at the end of the day, I have never worked for a company or even when I
326
00:20:57,563 --> 00:21:01,643
was on the partner side, worked with a company who listens to their customers
327
00:21:01,643 --> 00:21:05,543
as much as Snowflake does and makes decisions in that direction.
328
00:21:05,683 --> 00:21:08,763
Many people in the sales organization who have been here for a long time can
329
00:21:08,763 --> 00:21:12,723
point to specific features and then point to the specific customer who asked for it.
330
00:21:12,763 --> 00:21:16,223
And this goes back to things like Snowpipe. You know, data sharing came out
331
00:21:16,223 --> 00:21:18,683
of a very specific ask by one of our largest accounts.
332
00:21:19,103 --> 00:21:22,923
We want to be able to make this data available to our customers.
333
00:21:23,163 --> 00:21:27,703
This shift that you're seeing comes from a desire for our customers,
334
00:21:27,903 --> 00:21:33,243
for some of from some of our largest customers to consolidate from their tooling perspective, right?
335
00:21:33,343 --> 00:21:36,683
You know, you look at that slide that gets published every year of the tools
336
00:21:36,683 --> 00:21:40,003
of the modern data stack, and it just keeps getting bigger, right?
337
00:21:40,063 --> 00:21:41,723
It's just like every tool under
338
00:21:41,723 --> 00:21:44,563
the sun all of a sudden fits under this concept of modern data stack.
339
00:21:44,783 --> 00:21:49,783
And there was a scenario in sort of the early 2020s.
340
00:21:49,783 --> 00:21:52,983
So in the last kind of couple of years where everybody was buying all these
341
00:21:52,983 --> 00:21:57,683
like bespoke tools to do this job, and then it spend was running rampant,
342
00:21:57,683 --> 00:22:01,903
and then the pandemic hits, and then everybody's trying to like, save money.
343
00:22:02,003 --> 00:22:06,463
And they're like, wait a second, why do we own data IQ and data robot?
344
00:22:06,583 --> 00:22:10,203
What is happening here? This doesn't make any sense. So from a Snowflake perspective,
345
00:22:10,343 --> 00:22:13,463
right, for many of our largest customers who have consolidated and brought a
346
00:22:13,463 --> 00:22:14,983
lot of their data estate.
347
00:22:15,552 --> 00:22:19,372
To be stored and processed using Snowflake, they wanted to say,
348
00:22:19,472 --> 00:22:22,612
okay, wait, I want to be able to to code in Python, right?
349
00:22:22,672 --> 00:22:24,572
I want to be able to write data engineering pipelines in Python,
350
00:22:24,692 --> 00:22:25,752
here is Snowpark, right?
351
00:22:25,852 --> 00:22:28,932
So you have this ability now to code in the language that you feel more comfortable
352
00:22:28,932 --> 00:22:30,732
to build those declarative pipelines.
353
00:22:30,912 --> 00:22:35,412
But now my data science teams want to use Snowpark and, and we're not necessarily
354
00:22:35,412 --> 00:22:39,112
optimized, you know, our warehouses aren't necessarily optimized for those types of workloads.
355
00:22:39,312 --> 00:22:42,652
Okay, here comes Snowpark optimized warehouses. And then now we've We've been
356
00:22:42,652 --> 00:22:47,012
building applications. We love using Snowflake to power some of our analytical applications.
357
00:22:47,212 --> 00:22:51,192
We want to be putting more of our applications and consolidating this into Snowflake.
358
00:22:51,292 --> 00:22:52,492
Great. You guys build with containers.
359
00:22:52,912 --> 00:22:57,292
Awesome. Snowpark Container Services. It's very much a feedback loop with our customers.
360
00:22:57,632 --> 00:23:01,112
And yes, all of these things have the word snow in them. And we all have to be okay with that.
361
00:23:02,312 --> 00:23:06,832
They will always. And we're getting over it. Yeah. I think about 1999 moving
362
00:23:06,832 --> 00:23:11,132
forward in the on-prem world, let's say Oracle, right?
363
00:23:11,132 --> 00:23:16,272
The big challenge back then was that Oracle went off and bought all these companies
364
00:23:16,272 --> 00:23:17,412
and tried to integrate it.
365
00:23:17,532 --> 00:23:21,432
And as a consultant, I remember saying to them, yeah, they bought all this stuff,
366
00:23:21,612 --> 00:23:24,232
one tool, but you still have to self-integrate everything.
367
00:23:24,392 --> 00:23:27,452
It's painful, right? It's painful to integrate all these platforms.
368
00:23:27,832 --> 00:23:31,212
And what I find intriguing and interesting about where we are with cloud today
369
00:23:31,212 --> 00:23:36,792
is that everyone's pushing for at least the idea of these modular solutions
370
00:23:36,792 --> 00:23:40,992
where everyone has to have all these, the best. and I don't want vendor lock-in.
371
00:23:41,072 --> 00:23:42,512
I want to be able to swap things out.
372
00:23:42,712 --> 00:23:46,152
You're not going to do that. Like I look at it, I'm like, you're not going to do that.
373
00:23:47,012 --> 00:23:50,012
You're just not. Vendor lock-in is my least favorite term.
374
00:23:50,192 --> 00:23:54,572
Vendor lock-in is the funniest conversation I think I have because I'm just
375
00:23:54,572 --> 00:23:59,892
like, you're trying to say the problem we had in 2000 was vendor lock-in.
376
00:23:59,952 --> 00:24:00,892
That was not the problem.
377
00:24:00,972 --> 00:24:04,672
The problem was technology was not advanced enough to allow you to change.
378
00:24:04,912 --> 00:24:07,712
It's advanced enough now, regardless of what vendor you use,
379
00:24:07,752 --> 00:24:10,012
as long as it's cloud-based. It's actually pretty easy to move.
380
00:24:10,152 --> 00:24:14,472
I do love the idea of the consolidation that is starting to happen in the cloud
381
00:24:14,472 --> 00:24:17,852
where companies are getting smart about, all right, how do we do this?
382
00:24:17,932 --> 00:24:22,652
But they're trying to do it. Some do it well, smartly in terms of if I'm adding
383
00:24:22,652 --> 00:24:25,012
capabilities, it's built through our platform.
384
00:24:25,352 --> 00:24:29,252
Some people are acquiring tools and solutions and trying to integrate them in
385
00:24:29,252 --> 00:24:31,412
the cloud. And then it gets a little clunky and scary.
386
00:24:31,572 --> 00:24:35,812
We're now in a place where things are starting to get consolidated because of what you said.
387
00:24:35,992 --> 00:24:40,132
Look at the math. There's a billion little little dots on icons of companies
388
00:24:40,132 --> 00:24:44,652
on this giant capability map, and people are overwhelmed and confused.
389
00:24:44,912 --> 00:24:49,032
And I think Snowflake is and will always be a single SKU.
390
00:24:49,152 --> 00:24:52,952
When we make acquisitions, and when we build out net new features and capabilities.
391
00:24:53,432 --> 00:24:57,492
It's not something that you then have to turn around and pay for another license for.
392
00:24:57,592 --> 00:25:01,292
You buy your bucket of credits, and then you use them as you see fit.
393
00:25:01,412 --> 00:25:05,252
So back to your question about these decisions around LLMs and MLOps and and
394
00:25:05,252 --> 00:25:06,132
all that kind of stuff, right?
395
00:25:06,372 --> 00:25:10,072
It's it's for us, it's all about how are we driving that
396
00:25:10,511 --> 00:25:14,731
customer satisfaction in that NPS score, right? Like, we at Snowflake do not
397
00:25:14,731 --> 00:25:17,331
have a customer success org that does not exist at Snowflake.
398
00:25:17,471 --> 00:25:22,731
I didn't even, you know, I've worked with you guys since you worked with us years ago.
399
00:25:23,071 --> 00:25:27,971
And I didn't realize that this entire time. So that's yeah, yeah.
400
00:25:28,031 --> 00:25:32,851
And there's customer success office, like the idea of a customer success team
401
00:25:32,851 --> 00:25:34,911
not being there is support.
402
00:25:35,011 --> 00:25:38,411
Yes, success. No. And that's, that's an interesting differentiator,
403
00:25:38,471 --> 00:25:39,611
actually. so I appreciate that.
404
00:25:39,691 --> 00:25:44,171
I do want to pivot and give you the opportunity to continue the conversation
405
00:25:44,171 --> 00:25:46,511
we had over text regarding Gen AI.
406
00:25:46,811 --> 00:25:51,211
We have 15 minutes and I think it's important for us to cover this ground.
407
00:25:51,391 --> 00:25:55,591
When we were going back and forth, we were talking about the change in work, right?
408
00:25:55,671 --> 00:26:01,211
How is Gen AI actually going to impact work and maybe not just work for the
409
00:26:01,211 --> 00:26:03,831
general masses, but work for data teams?
410
00:26:04,031 --> 00:26:07,171
Yeah, so I want to go there with you. I'm curious of your opinion.
411
00:26:07,171 --> 00:26:12,691
I'm reading back through our texts to make sure I'm tracking to what I say.
412
00:26:13,051 --> 00:26:17,171
I think you've made a comment because people think that they're going to lose their jobs.
413
00:26:17,411 --> 00:26:22,311
Yes. But this is also what people who had tied their whole life to Oracle or
414
00:26:22,311 --> 00:26:27,911
Teradata, the DBAs of the world, felt when people started moving to the cloud.
415
00:26:28,251 --> 00:26:31,811
They pushed back really hard because they felt threatened. It is okay.
416
00:26:31,931 --> 00:26:33,491
It is okay to feel threatened.
417
00:26:33,651 --> 00:26:36,771
It is okay to feel like your job might be at risk. it's
418
00:26:36,771 --> 00:26:39,691
not okay to stop innovation because you don't want
419
00:26:39,691 --> 00:26:42,391
to change yourself right i think the the best advice i could
420
00:26:42,391 --> 00:26:45,231
give to to anybody in this space at this point is to be
421
00:26:45,231 --> 00:26:48,591
flexible in my own career right i started as a visualization guy
422
00:26:48,591 --> 00:26:51,491
i was a front-end bi analyst kind of person
423
00:26:51,491 --> 00:26:56,811
right and as the cloud became more prevalent the the bi end of the spectrum
424
00:26:56,811 --> 00:27:02,811
became less relevant and for me it it became important to diversify my skillset
425
00:27:02,811 --> 00:27:10,031
and be able to get an understanding of how ultimately do these BI products get fed with data?
426
00:27:10,367 --> 00:27:15,167
So I would say to anybody who is feeling threatened by generative AI that,
427
00:27:15,227 --> 00:27:17,987
but somebody's got to build the model. Somebody's got to tune the model.
428
00:27:18,207 --> 00:27:22,487
Somebody has to deploy the model. Somebody has to own the application that then
429
00:27:22,487 --> 00:27:24,427
the users interact with.
430
00:27:24,547 --> 00:27:27,227
If you feel like your job is going to be eliminated because you're not building
431
00:27:27,227 --> 00:27:33,227
ETL pipelines and you're not building dashboards anymore, then go take a course on LLMs.
432
00:27:33,287 --> 00:27:37,147
Go take a course on GPUs. Like, I understand that these are extremely technical
433
00:27:37,147 --> 00:27:40,667
constructs. And so it takes a lot to learn about them.
434
00:27:40,747 --> 00:27:45,827
If your other option is to just be a curmudgeon and not let innovation happen,
435
00:27:46,007 --> 00:27:50,847
because you're worried about your job, that's not ideal either.
436
00:27:50,947 --> 00:27:54,367
This these things will happen whether you get in the way or not.
437
00:27:54,367 --> 00:27:56,567
I'm going to print t-shirts curmudgeon.
438
00:27:57,007 --> 00:27:59,827
I totally agree with you.
439
00:27:59,907 --> 00:28:07,067
And I kind of mentioned this right during the first episode when I was a front
440
00:28:07,067 --> 00:28:11,547
end developer, I built BI solutions from scratch.
441
00:28:11,547 --> 00:28:14,587
And the pain and agony of
442
00:28:14,587 --> 00:28:17,827
seeing a what you see is what you get drag and
443
00:28:17,827 --> 00:28:21,887
drop lizzy wig bi platform in 2003
444
00:28:21,887 --> 00:28:28,067
i recalls when i first saw it from oracle yeah i almost had a heart attack because
445
00:28:28,067 --> 00:28:35,367
obi it was pre-obi it's kind of like it was it was the precursor to obi and
446
00:28:35,367 --> 00:28:39,087
i i still remember that i almost had a heart attack i didn't let anybody know
447
00:28:39,087 --> 00:28:41,687
that i was scared out of my mind and i I pivoted,
448
00:28:41,687 --> 00:28:47,487
I think the thing here is for people to realize that somebody has to design these solutions,
449
00:28:48,240 --> 00:28:51,340
you have to understand how these things are going to work. And you have to understand
450
00:28:51,340 --> 00:28:53,040
data quality is part of the problem.
451
00:28:53,360 --> 00:28:56,900
All the issues surrounding the work we do are still going to exist.
452
00:28:57,260 --> 00:29:00,480
It's just how we solve those issues are going to change. And you have to get
453
00:29:00,480 --> 00:29:03,900
uncomfortable more often now than we did in the past.
454
00:29:04,140 --> 00:29:10,740
I had a customer, he basically was like, so can I just train a model and get rid of ETL entirely?
455
00:29:11,280 --> 00:29:14,500
And I was like, no. I was like, Like in theory, sure.
456
00:29:14,700 --> 00:29:18,020
But is the cost of training that model going to be so great?
457
00:29:18,120 --> 00:29:21,420
And oh, by the way, is it going to perform like crap? Yes.
458
00:29:21,600 --> 00:29:24,720
So you're not going to get rid of ETL. And in the same way that,
459
00:29:24,720 --> 00:29:27,900
you know, Sandy, you and I had this debate over text about dashboards, right?
460
00:29:28,060 --> 00:29:33,720
I understand that the chat interface will make it easier for users to interact
461
00:29:33,720 --> 00:29:36,380
with the data in a way that is more exploratory.
462
00:29:36,600 --> 00:29:40,780
But what that's really going to get rid of is more of the Excel type of model
463
00:29:40,780 --> 00:29:44,480
of exploring and slicing and dicing data in certain ways.
464
00:29:44,600 --> 00:29:47,960
I pray to God that that happens. Your lips to God's ears, honestly.
465
00:29:48,040 --> 00:29:52,020
But there is still going to be a space for dashboard.
466
00:29:52,200 --> 00:29:54,660
It's not going to eliminate other ways of interacting with data,
467
00:29:54,760 --> 00:29:58,700
except for obviously in the more of the exploratory style, right?
468
00:29:58,780 --> 00:30:01,460
You're going to get a lot more people who are going to ask questions of data
469
00:30:01,460 --> 00:30:06,300
themselves and not go ask a BI analyst to go pull a report for them to answer
470
00:30:06,300 --> 00:30:08,720
one specific question. I think those days are gone.
471
00:30:08,940 --> 00:30:11,480
I think those days are, they need to be gone.
472
00:30:11,960 --> 00:30:17,240
I want everybody to move towards decision intelligence, where you're actually
473
00:30:17,240 --> 00:30:21,820
supporting decisions people have to make in the workflow in which they're making them.
474
00:30:22,300 --> 00:30:27,780
That's where I think this all needs to go towards. And the days of I go to a
475
00:30:27,780 --> 00:30:30,220
dashboard need to go away. way.
476
00:30:30,280 --> 00:30:34,480
I spent probably the first 10 years of my career doing nothing but front-end
477
00:30:34,480 --> 00:30:37,780
solutioning. And all I could think about is getting rid of it now.
478
00:30:37,940 --> 00:30:41,340
I'm just like, I'm so sick of it. I just want to help people make the right
479
00:30:41,340 --> 00:30:44,600
decision at the right time, in the right workflow, et cetera.
480
00:30:44,800 --> 00:30:47,240
I think that needs to start happening sooner rather than later,
481
00:30:47,300 --> 00:30:52,340
but there's so many barriers to that because BI teams are still working in isolation
482
00:30:52,340 --> 00:30:55,100
from the AI team, from the digital team.
483
00:30:55,360 --> 00:31:02,140
I think I understand where you're coming from, But I think that the dashboard
484
00:31:02,140 --> 00:31:04,500
paradigm is never going to go away, right?
485
00:31:04,620 --> 00:31:09,100
And even if the dashboard paradigm shifts to being basically,
486
00:31:09,160 --> 00:31:11,520
how do I embed KPIs into your workflow?
487
00:31:11,740 --> 00:31:15,740
That still has a dashboard-like feel, right?
488
00:31:15,900 --> 00:31:18,880
And part of our debate over text was around operational, right?
489
00:31:18,940 --> 00:31:21,420
The operational dashboards will continue to exist, period.
490
00:31:21,740 --> 00:31:26,820
That's just how it's going to work, right? So I get decision intelligence and
491
00:31:26,820 --> 00:31:32,000
I get making that point. But like, when it comes to a user interacting with
492
00:31:32,000 --> 00:31:36,180
data, you have to make sure that it fits in terms of their own maturity.
493
00:31:37,429 --> 00:31:43,089
Putting it all in chatbots is good, except for the users who don't know how
494
00:31:43,089 --> 00:31:47,009
to ask questions, who just want the specific answer to the specific question
495
00:31:47,009 --> 00:31:50,289
that they ask themselves every day to do their job. It's that bad.
496
00:31:51,129 --> 00:31:56,269
You don't know how to ask questions? Wow. I mean, come on.
497
00:31:56,309 --> 00:32:01,229
You and I have spent enough time with enough clients in this world who ask the worst questions.
498
00:32:02,249 --> 00:32:07,129
Let's be clear. here. I plead the fifth on that one. I am not subscribing to that, James.
499
00:32:07,489 --> 00:32:10,329
Well, you still have clients to deal with. I get to be off the center side.
500
00:32:10,449 --> 00:32:12,969
I don't have to do that. There are no dumb questions.
501
00:32:13,689 --> 00:32:15,809
Well, I do tell people there's no such thing as a dumb question,
502
00:32:15,869 --> 00:32:19,069
only a dumb answer. But it doesn't mean you don't. It doesn't mean you know
503
00:32:19,069 --> 00:32:20,049
how to ask the question, right?
504
00:32:20,249 --> 00:32:24,309
I mean, that applies to to generative AI as well, right?
505
00:32:24,409 --> 00:32:29,029
I mean, there are dumb answers that come back from from these models as well.
506
00:32:29,089 --> 00:32:31,849
We have to be able to one, except you're going
507
00:32:31,849 --> 00:32:34,749
to get a dumb answer but to put guardrails in
508
00:32:34,749 --> 00:32:37,569
place the human in the loop to to make
509
00:32:37,569 --> 00:32:40,729
the decision around whether or not that's the answer we can actually use
510
00:32:40,729 --> 00:32:46,009
and trust yeah i could not agree with that statement more like i as much as
511
00:32:46,009 --> 00:32:49,769
as much as ml and this idea of prescriptive analytics in terms of like actually
512
00:32:49,769 --> 00:32:53,729
letting the computer make decisions i like there's a world where i understand
513
00:32:53,729 --> 00:33:00,269
why that that that happens so like google maps right nobody's in nobody is sitting at the terminal.
514
00:33:00,989 --> 00:33:03,949
Seeing all the google maps requests in and then telling them
515
00:33:03,949 --> 00:33:06,709
which directions to go right a
516
00:33:06,709 --> 00:33:09,849
computer is making that decision i i get that and guess
517
00:33:09,849 --> 00:33:12,749
what google maps does it wrong a lot and so
518
00:33:12,749 --> 00:33:16,129
you have to you as the user has to decide i trust
519
00:33:16,129 --> 00:33:18,949
this or no i really don't so we were we
520
00:33:18,949 --> 00:33:21,749
went up to north my my family and i went up to northern maine for
521
00:33:21,749 --> 00:33:24,789
the eclipse last week it was beautiful it was it
522
00:33:24,789 --> 00:33:27,469
was a beautiful day it was the coolest experience i've ever had but
523
00:33:27,469 --> 00:33:30,929
it took us what should have been a at most five
524
00:33:30,929 --> 00:33:33,589
hour drive took eight hours to get home google maps just like
525
00:33:33,589 --> 00:33:36,569
didn't work and everybody ended up on the the area at
526
00:33:36,569 --> 00:33:39,489
franconia where 93 and three come together and they went
527
00:33:39,489 --> 00:33:42,429
five miles in five hours i i i
528
00:33:42,429 --> 00:33:45,629
agree i think it was we can find instances where
529
00:33:45,629 --> 00:33:48,789
it's not going to work when there's an event that is so mass
530
00:33:48,789 --> 00:33:51,469
in terms of the number of people jumping on the highway and the
531
00:33:51,469 --> 00:33:54,509
information doesn't get back fast enough for it to update its
532
00:33:54,509 --> 00:33:57,129
algorithm yeah of course but i
533
00:33:57,129 --> 00:33:59,969
but i i looked at the map and i said i don't
534
00:33:59,969 --> 00:34:04,789
like what this route is and i had to go manually put in a point on the map to
535
00:34:04,789 --> 00:34:10,609
force google maps to take me in that direction right so like that's that's sort
536
00:34:10,609 --> 00:34:14,369
of the manual intervention that you're talking about anjali of like i as a user
537
00:34:14,369 --> 00:34:18,869
have to be smart enough and say this doesn't make any sense Generally,
538
00:34:18,889 --> 00:34:22,869
I might get rid of some level of IT jobs in theory.
539
00:34:22,989 --> 00:34:25,569
Yeah, I don't think the jobs are going to go away. I think they're just not
540
00:34:25,569 --> 00:34:28,709
going to scale the way they've been scaling. Sure.
541
00:34:29,029 --> 00:34:31,669
That's kind of where I'm at. The scalability is going down a bit.
542
00:34:31,669 --> 00:34:35,909
Like you don't have to have thousands of data engineers if you're a data heavy company.
543
00:34:36,321 --> 00:34:39,481
That's going to probably simmer down a bit. You'll be able to do a lot more
544
00:34:39,481 --> 00:34:42,961
with less. So I don't see anything completely disappearing.
545
00:34:43,221 --> 00:34:49,921
However, I will say the logo for their podcast, a logo for a friend's Facebook group.
546
00:34:50,001 --> 00:34:53,221
She was pinging me yesterday, just two days ago, asking for a logo.
547
00:34:53,441 --> 00:34:57,021
Hey, you did this AI thing for your podcast logo. Can you go do one for me?
548
00:34:57,201 --> 00:35:00,041
And it was like around the barrel, I think it's called. She's like,
549
00:35:00,061 --> 00:35:01,901
it's a new Facebook, but I'm trying to do it. I was like, fine.
550
00:35:01,961 --> 00:35:03,901
So I literally opened up my phone.
551
00:35:04,001 --> 00:35:06,441
I didn't even use my computer. I had chat GPT on my phone.
552
00:35:06,581 --> 00:35:11,061
I asked it and within the third hit, it had it. Five minutes later, she's got a logo.
553
00:35:11,301 --> 00:35:16,221
And so I feel bad for all the artists out there on Fiverr that was creating
554
00:35:16,221 --> 00:35:19,901
logos for five bucks or a hundred bucks a pop. That industry is gone.
555
00:35:20,081 --> 00:35:26,601
So apologies to them. But I think that's going to be the tough one for us.
556
00:35:26,781 --> 00:35:31,081
I do have to wrap this up. I have a client call that I had scheduled after this,
557
00:35:31,101 --> 00:35:32,521
unfortunately, so I got to run over there.
558
00:35:32,601 --> 00:35:36,721
But James, James, what is your last piece of advice to anybody who's scared
559
00:35:36,721 --> 00:35:40,181
to death in terms of, I'm a data engineer, I'm just starting my career.
560
00:35:40,381 --> 00:35:42,161
What should I focus on? Learn Python.
561
00:35:42,461 --> 00:35:45,441
There's a lot of people who will tell you, oh no, SQL will never go away.
562
00:35:45,581 --> 00:35:46,861
Yeah. They're not correct.
563
00:35:47,141 --> 00:35:51,961
Python is, from a declarative engineering perspective, Python is where people are moving.
564
00:35:52,361 --> 00:35:56,441
And Python is super flexible. I said earlier, being flexible is the most important
565
00:35:56,441 --> 00:35:57,421
thing that you can do for yourself.
566
00:35:57,661 --> 00:36:00,741
And I think, at least in terms of what's happening in the next five years,
567
00:36:00,981 --> 00:36:05,641
having an understanding of Python is is going to allow you to do a lot of things
568
00:36:05,641 --> 00:36:07,381
and be and help be successful.
569
00:36:07,621 --> 00:36:12,141
So if I had any advice for a budding data engineer, the first thing I would
570
00:36:12,141 --> 00:36:14,181
tell you is to go learn Python, learn Python.
571
00:36:14,421 --> 00:36:16,881
I love it. Well, thank you, James, for being on.
572
00:36:17,181 --> 00:36:20,861
Of course. Thank you for having me. I mean, it was exactly what I expected.
573
00:36:20,981 --> 00:36:23,221
I expected this from you. So I appreciate that.
574
00:36:23,361 --> 00:36:27,821
And we probably need to have a round two in 10 years to figure out was I right
575
00:36:27,821 --> 00:36:29,481
about operational dashboards?
576
00:36:29,561 --> 00:36:32,021
Or was James's right about operational dashboards?
577
00:36:32,161 --> 00:36:35,981
I'm putting the that down now. I'll Venmo the money to you and let's do it.
578
00:36:36,041 --> 00:36:40,141
Done. Thank you. Well, have a great weekend. Bye, Anjali. I'm off.
579
00:36:40,400 --> 00:37:08,668
Music.