1
00:00:00,160 --> 00:00:03,939
Welcome to Data Driven, where we dive into the thrilling world of data,

2
00:00:04,080 --> 00:00:07,839
AI, and on occasion, misbehaving chatbots suggesting

3
00:00:07,839 --> 00:00:11,300
glue for your pizza. This episode features Bar Moses,

4
00:00:11,599 --> 00:00:15,219
CEO of Monte Carlo. Not the casino, not the car,

5
00:00:15,225 --> 00:00:18,685
but the company keeping your data from quietly wrecking your business.

6
00:00:19,065 --> 00:00:22,605
We talk observability, the chaos of unreliable data,

7
00:00:22,745 --> 00:00:25,465
and why one tiny schema change cost a company

8
00:00:25,465 --> 00:00:28,904
$100,000,000. Ouch. So buckle

9
00:00:28,904 --> 00:00:32,530
up. Because if your AI bots are making decisions without

10
00:00:32,530 --> 00:00:35,970
reliable data, well, hope you like eating rocks for the

11
00:00:35,970 --> 00:00:39,809
minerals. Hello, and

12
00:00:39,809 --> 00:00:42,950
welcome back to Data Driven, the podcast where we explore the emergent

13
00:00:43,485 --> 00:00:47,004
fields of data science, artificial intelligence, and, of course, data

14
00:00:47,004 --> 00:00:50,605
engineering. And with me today is my favorite data engineer in the

15
00:00:50,605 --> 00:00:54,364
world, Andy Leonard. How's it going, Andy? It's going well, Frank.

16
00:00:54,364 --> 00:00:57,820
How are you? I'm doing well. I'm doing well. I was in Raleigh last

17
00:00:57,820 --> 00:01:01,440
week, drove down, rented a car actually,

18
00:01:02,539 --> 00:01:06,240
to save mileage on, on ours, and,

19
00:01:07,180 --> 00:01:10,159
spoiled because it's been a while since I bought a new car. And

20
00:01:10,955 --> 00:01:13,675
this is the second time I rented a car, and I'm getting tempted. I ain't

21
00:01:13,675 --> 00:01:17,515
getting tempted. It was a Chevy. It was

22
00:01:17,515 --> 00:01:21,295
a Chevy Malibu. Not a Monte not a Monte Carlo.

23
00:01:21,755 --> 00:01:25,210
See what I did there? I don't even know if they still make them. I

24
00:01:25,210 --> 00:01:28,890
I was driving, the little one off and dropping the little one off at daycare,

25
00:01:28,890 --> 00:01:32,650
and I was behind a Chevy Monte Carlo, like, a two early

26
00:01:32,650 --> 00:01:36,190
two thousands vintage. But that is actually quite relevant

27
00:01:36,275 --> 00:01:40,115
to our discussion today because with us today, we have Bar Moses, who is the

28
00:01:40,115 --> 00:01:43,715
CEO and cofounder of Monte Carlo, the data

29
00:01:43,715 --> 00:01:47,415
and AI reliability company, not the casino

30
00:01:47,475 --> 00:01:51,130
or the car, I would assume, or the town. Monte Carlo

31
00:01:51,130 --> 00:01:54,729
is the creator of the industry's first end to end data and

32
00:01:54,729 --> 00:01:58,250
AI, observability platform with

33
00:01:58,250 --> 00:02:01,630
$236,000,000 in funding from Accel

34
00:02:01,690 --> 00:02:05,435
Iconic Growth and others. They are on a mission to bring

35
00:02:05,435 --> 00:02:08,955
trustworthy and reliable data and AI, to

36
00:02:08,955 --> 00:02:12,395
companies everywhere. The company was recently recognized as

37
00:02:12,395 --> 00:02:15,855
a enterprise tech 30 company, a CRN

38
00:02:16,155 --> 00:02:19,280
emerging vendor, and an inc.com,

39
00:02:19,740 --> 00:02:23,120
best workplace and accounts Fox, Roche,

40
00:02:23,340 --> 00:02:27,180
Nasdaq, and PagerDuty, among others, as their customers. Welcome

41
00:02:27,180 --> 00:02:30,620
to the show, Bar. Thank you so much. Great to be here, Frank and

42
00:02:30,620 --> 00:02:34,424
Andy. Awesome. An intro. No problem. Do you drive a

43
00:02:34,424 --> 00:02:38,185
Monte Carlo? Because that would be epic. You know, I really should

44
00:02:38,185 --> 00:02:42,025
be driving a Monte Carlo. I do not, and I've never actually been to

45
00:02:42,025 --> 00:02:45,864
Monte Carlo either. So I will tell you if you're into cars,

46
00:02:45,864 --> 00:02:49,569
like, I'm like a recovering car, nerd. Oh,

47
00:02:49,569 --> 00:02:53,250
very cool. It looks like a car show. Like, honestly, I went to Monte

48
00:02:53,250 --> 00:02:57,010
Carlo, and we had rented, like, a Saab convertible. And I felt like we were

49
00:02:57,010 --> 00:03:00,769
driving. We were driving driving, like, the low end

50
00:03:00,769 --> 00:03:04,565
of the car thing. I mean, there were I mean, I've never

51
00:03:04,565 --> 00:03:08,405
seen Bentleys in the wild, like, just parked on the street,

52
00:03:08,405 --> 00:03:11,924
like, no big deal. Wow. Like, I mean, every

53
00:03:11,924 --> 00:03:15,525
luxury car if you're in a Saab and you feel like you're slumming it

54
00:03:15,685 --> 00:03:18,300
Yeah. It is clearly a high money area.

55
00:03:20,040 --> 00:03:23,879
But, so welcome to the show. So Monte Carlo

56
00:03:23,959 --> 00:03:26,599
why'd you get the name? I I'm assuming it might have something to do with

57
00:03:26,599 --> 00:03:30,120
Monte Carlo simulations, but that's in the Great question. Yeah. The

58
00:03:30,120 --> 00:03:33,885
unofficial story is that, one of our CO, founders is a fan

59
00:03:33,885 --> 00:03:36,945
of formula one and, you know, as, you know, formula one crisis.

60
00:03:37,805 --> 00:03:41,405
So right. That's, you know, clearly the, the, that's the

61
00:03:41,405 --> 00:03:45,120
unofficial story. The official story is that, you know, we

62
00:03:45,120 --> 00:03:48,480
had to we had to name the company. We started working with customers when we

63
00:03:48,480 --> 00:03:51,540
started the company, and we we had to choose some name.

64
00:03:52,319 --> 00:03:56,160
And, I studied math and stats in college, and so I sort

65
00:03:56,160 --> 00:03:59,385
of opened my my stats book and sort of looked through and,

66
00:03:59,864 --> 00:04:02,845
you know, reviewed my option and, you know, Markov,

67
00:04:03,704 --> 00:04:07,545
chains didn't seem like a great name. And next up was

68
00:04:07,545 --> 00:04:11,305
Bayes' theorem, which was similarly kind of not great. And

69
00:04:11,305 --> 00:04:14,602
and then, you know, I was reminded of Monte Carlo and Monte Carlo simulations. I

70
00:04:14,602 --> 00:04:17,519
actually I actually did some work with Monte Carlo simulations earlier in my career.

71
00:04:18,620 --> 00:04:21,419
And it seemed like it seemed like a great name, a name that would speak

72
00:04:21,419 --> 00:04:25,260
to, you know, data engineers, data analysts, folks that have been the space.

73
00:04:25,260 --> 00:04:29,074
And, you know, I think naming a company is a very difficult

74
00:04:29,074 --> 00:04:32,354
thing to do. We decided to go with it. And the spirit of Monte Carlo,

75
00:04:32,354 --> 00:04:36,115
One of our values is ship and iterate. And so, the

76
00:04:36,115 --> 00:04:39,875
name has sort of stuck with us since. And, it's quite memorable. People either

77
00:04:39,875 --> 00:04:43,610
love it or hate it. So I think it works for us. I think it

78
00:04:43,610 --> 00:04:46,490
it works. Like, I think of the car. I think of the casinos. It has

79
00:04:46,490 --> 00:04:50,090
a certain amount of, high class, maybe more so than Markov

80
00:04:50,090 --> 00:04:53,850
chains, Markov chains. Although I did for a time flirt with the

81
00:04:53,850 --> 00:04:57,575
idea of of also starting a company called Markoff Chains, but,

82
00:04:57,575 --> 00:05:00,935
like, have see if we could see if we can get money for mister t

83
00:05:00,935 --> 00:05:04,775
to be the spokesman. That would

84
00:05:04,775 --> 00:05:08,455
have been epic. Yeah. Jeez. He did you. Ideas, Fran. I was the

85
00:05:08,455 --> 00:05:11,360
only one I was the only one that thought that was a good idea, but,

86
00:05:11,360 --> 00:05:14,900
you know, I was a big fan of mister t as a kid. Marketing. Yeah.

87
00:05:14,960 --> 00:05:17,860
That's funny. That's what I do in my day job now. Oh, yeah.

88
00:05:20,479 --> 00:05:22,485
I swear, folks, I didn't pay her to say that.

89
00:05:24,564 --> 00:05:28,164
So so you you talk about data and I AI

90
00:05:28,164 --> 00:05:31,305
reliability. And to me, when when I hear that,

91
00:05:32,245 --> 00:05:35,625
a slew of things come to mind. Like, there's security, there's the

92
00:05:36,599 --> 00:05:40,120
veracity, like, the five v's and all that or four v's or whatever it

93
00:05:40,120 --> 00:05:43,960
was. What exactly is kind of Monte Carlo's, like,

94
00:05:43,960 --> 00:05:47,639
wheelhouse there? Yeah. Great question. I'll

95
00:05:47,639 --> 00:05:50,895
actually sort of anchor ourselves in in kind of the metaphor or sort of a

96
00:05:50,974 --> 00:05:54,754
corollary that we like to use here, which is really based on software engineering.

97
00:05:54,974 --> 00:05:58,435
So we didn't reinvent the wheel when we say data and AI observability.

98
00:05:59,294 --> 00:06:02,595
We really take concepts that work for engineering and adapt them.

99
00:06:03,550 --> 00:06:07,070
So, you know, when we started the company, the idea, the

100
00:06:07,070 --> 00:06:10,910
hypothesis, the the thesis that we started the company on was data

101
00:06:10,910 --> 00:06:14,430
was going to be as important to businesses as applications, as online

102
00:06:14,430 --> 00:06:18,264
applications. And, they were data was going to

103
00:06:18,264 --> 00:06:22,105
drive the most critical sort of, you know, lifeblood of companies through

104
00:06:22,105 --> 00:06:25,324
decision making, internal products, external products.

105
00:06:25,945 --> 00:06:29,785
And, while software engineers had all the solutions and tools in the

106
00:06:29,785 --> 00:06:33,430
world to make sure their applications were reliable, and so some, you

107
00:06:33,730 --> 00:06:37,570
know, some off the shelf solutions like Datadog, New Relic, Splunk might be

108
00:06:37,570 --> 00:06:41,090
familiar to you, data teams were flying wide. So there was literally

109
00:06:41,090 --> 00:06:44,865
nothing that they could use to know that their data was

110
00:06:44,865 --> 00:06:48,625
actually accurate and trusted. That's sort of, like, the the problem the core problem that

111
00:06:48,625 --> 00:06:52,245
we started. Fast forward to today, you know, we created the data observability

112
00:06:52,465 --> 00:06:56,065
category. We're continuing to create it. AI is making this problem just

113
00:06:56,065 --> 00:06:59,880
infinitely bigger, harder, more important. Why? Because

114
00:06:59,880 --> 00:07:03,340
data and AI products are now you know, there's a proliferation of those.

115
00:07:04,680 --> 00:07:08,380
An AI application is only as good as the data that's powering it,

116
00:07:08,600 --> 00:07:12,145
and the AI application itself can be inaccurate, can be

117
00:07:12,145 --> 00:07:15,824
unreliable. Right? And so at a very high level

118
00:07:15,985 --> 00:07:19,664
I know this is, you know, very vague, but at a very high

119
00:07:19,664 --> 00:07:23,104
level, the idea was the same diligence that we treat software

120
00:07:23,104 --> 00:07:26,830
applications, we should be treating for data and AI applications. Now,

121
00:07:26,830 --> 00:07:30,350
what does that actually mean? How do we do that? Enter the concept of

122
00:07:30,350 --> 00:07:33,790
observability. Observability is basically understanding or

123
00:07:33,790 --> 00:07:36,690
assessing a system's health based on its output.

124
00:07:37,870 --> 00:07:41,455
And so basically, the thesis was, can we observe end to end the

125
00:07:41,455 --> 00:07:45,135
data and AI estate, learn what the patterns

126
00:07:45,135 --> 00:07:48,915
are in the in the data, bring together metadata and context,

127
00:07:49,055 --> 00:07:52,675
lineage, for example, about the data, derive insights

128
00:07:52,735 --> 00:07:56,409
based on that to understand and determine what the system should

129
00:07:56,409 --> 00:08:00,169
behave like, and alert if that gets violated. So that's sort

130
00:08:00,169 --> 00:08:03,930
of the first part. The first is actually being being able to help data teams

131
00:08:03,930 --> 00:08:07,245
detect issues. The second part is actually being help,

132
00:08:07,405 --> 00:08:11,245
helping data teams resolve issues. Now here's the interesting thing

133
00:08:11,245 --> 00:08:14,764
that we sort of learned over over the years. We've worked with hundreds of of

134
00:08:14,764 --> 00:08:18,285
enterprises. So, you know, we mentioned a few. We real really work with the top

135
00:08:18,285 --> 00:08:20,685
companies in every single industry. So,

136
00:08:22,009 --> 00:08:25,550
you know, in in, in health care, in retail,

137
00:08:26,009 --> 00:08:29,690
in manufacturing, in, technology, in each of these

138
00:08:29,690 --> 00:08:33,289
areas, the, data in the state

139
00:08:33,289 --> 00:08:36,925
obviously varies, but there are actually interestingly commonalities. And the

140
00:08:36,925 --> 00:08:40,525
commonalities is that every single issue can be

141
00:08:40,525 --> 00:08:44,145
traced back to a problem with the data, problem with the code,

142
00:08:44,685 --> 00:08:48,445
problem with the system, or problem with the model output. Can go

143
00:08:48,445 --> 00:08:51,565
into detail into more each of those, but that's sort of the high level,

144
00:08:51,885 --> 00:08:55,560
framework. We basically provide end to end coverage to help data teams

145
00:08:55,560 --> 00:08:59,400
understand what the issues are and help them trace them back to data issues,

146
00:08:59,400 --> 00:09:03,080
code issues, system issues, or model output issues. So when did

147
00:09:03,080 --> 00:09:06,825
you get the idea that I'm sorry, Andy. I cut you off. Okay. When

148
00:09:06,985 --> 00:09:10,125
did you get the idea when you realized that data is gonna be as important

149
00:09:10,185 --> 00:09:13,805
as applications are to businesses? Oh, great question.

150
00:09:13,945 --> 00:09:17,325
Yeah. Great question. So so we started the company in 2019.

151
00:09:17,705 --> 00:09:21,465
And, actually, what's interesting, it was pretty clear to us then, but we

152
00:09:21,465 --> 00:09:25,270
had to prove that or we had to convince that of people. Definitely.

153
00:09:25,810 --> 00:09:29,410
Yeah. It was not obvious. It's it's still there's still a

154
00:09:29,410 --> 00:09:33,010
lot of people that are kind of, like, I guess, they'd be in the quadrant

155
00:09:33,010 --> 00:09:35,750
of laggards where they realize, oh, I guess this is important.

156
00:09:36,764 --> 00:09:40,605
A %. I would imagine in 2019, that would have

157
00:09:40,605 --> 00:09:44,125
been you would have sounded insane. Like We we sound I

158
00:09:44,125 --> 00:09:47,884
sounded insane a %. People are like, what? Data is

159
00:09:47,884 --> 00:09:51,680
gonna be important? Are you sure? Now a couple of things happened

160
00:09:51,680 --> 00:09:54,580
since, which I think helped. First is,

161
00:09:55,600 --> 00:09:59,120
there were some large acquisitions in the data space, like Tableau and

162
00:09:59,120 --> 00:10:02,800
Looker earlier on, and then Snowflake IPO'd. Snowflake was the

163
00:10:02,800 --> 00:10:06,315
largest software IPO of all times. It was quite interesting that the

164
00:10:06,315 --> 00:10:09,995
largest software IPO of all time is a data company. So I think those

165
00:10:09,995 --> 00:10:13,695
things sort of help kind of convince that this you know,

166
00:10:13,755 --> 00:10:17,130
convince, at least, externally, you know,

167
00:10:17,530 --> 00:10:21,050
to the market that data will continue to be will will be

168
00:10:21,050 --> 00:10:24,730
important and critical. I think the things that I noticed is, you know,

169
00:10:24,730 --> 00:10:28,010
before we even started the company, we spoke to hundreds of data leaders, and I

170
00:10:28,010 --> 00:10:31,725
speak to dozens of data leaders every single month. They continue

171
00:10:32,185 --> 00:10:35,325
and I think what you hear from them is more and more

172
00:10:35,945 --> 00:10:39,705
data teams and software engineering teams are building products hand in hand.

173
00:10:39,705 --> 00:10:43,085
So they're actually they're side by side building. Right? And so, actually,

174
00:10:43,305 --> 00:10:46,365
almost more and more critical business

175
00:10:47,279 --> 00:10:51,120
applications, revenue generating products are based off of

176
00:10:51,120 --> 00:10:54,959
data, and they're being powered by data. I'm not even talking

177
00:10:54,959 --> 00:10:58,800
about generative AI, which is a whole whole other story why that matters, but just

178
00:10:58,800 --> 00:11:02,180
data products by itself. Think about reports that people look at internally.

179
00:11:02,634 --> 00:11:06,315
You know, just give you an example. You know, we work with with, many

180
00:11:06,315 --> 00:11:09,834
airlines, for example. Airlines have a lot of data that goes to internal

181
00:11:09,834 --> 00:11:13,675
operations. Like, what's the connecting flight? What's your flight number? How

182
00:11:13,675 --> 00:11:17,274
many flights left today? What time did they leave? How many passengers were on

183
00:11:17,274 --> 00:11:20,850
the airplane? Where is your luggage? Right? That

184
00:11:20,850 --> 00:11:24,690
information is powering internal and external products. You know, it's powering the application

185
00:11:24,690 --> 00:11:28,449
that you're using in order to onboard the the plane, in order to connect

186
00:11:28,449 --> 00:11:31,589
to your next flight. If that data is inaccurate, like,

187
00:11:32,575 --> 00:11:36,335
you're screwed. Right? And that hurts tremendously. Your brand

188
00:11:36,335 --> 00:11:39,855
is an as an airline, your reputation, it leads to

189
00:11:39,855 --> 00:11:43,215
reduced revenue, increased regulatory risk that you're putting

190
00:11:43,215 --> 00:11:46,675
yourself. Right? So so the data,

191
00:11:47,070 --> 00:11:50,830
what we see from our customers is powering critical use cases like

192
00:11:50,830 --> 00:11:54,327
airlines. I'll give you another example. You know, we work with a,

193
00:11:54,635 --> 00:11:58,350
you know, a Fortune 500 company, perhaps your your favorite cereal.

194
00:11:58,350 --> 00:12:01,630
I don't know if you're you guys are big cereal. I I, like, eat cereal

195
00:12:01,630 --> 00:12:04,505
for breakfast, lunch, and and dinner. It's, like, my go to.

196
00:12:05,285 --> 00:12:09,125
You'd be surprised into how much data optimization, machine learning,

197
00:12:09,125 --> 00:12:12,565
and AI goes into actually optimizing the number and

198
00:12:12,565 --> 00:12:16,170
location of cereal on the shelf. So there's a lot of

199
00:12:16,170 --> 00:12:19,930
data that goes into supply chain management to make sure that you're

200
00:12:19,930 --> 00:12:22,029
actually, like, fulfilling the right warehouse,

201
00:12:24,009 --> 00:12:27,850
demands on time and, you know, making sure that everyone gets

202
00:12:27,850 --> 00:12:31,165
their serial on time. There's actually a lot of data that goes into all of

203
00:12:31,165 --> 00:12:35,005
that. So I think what gave me conviction was in speaking with

204
00:12:35,005 --> 00:12:38,765
so many companies across so many industries, data was

205
00:12:38,765 --> 00:12:41,905
actually allowing data teams, allowing

206
00:12:42,285 --> 00:12:45,780
organizations to build better products, to build more

207
00:12:45,840 --> 00:12:49,220
personalized products, and to make better decisions about the organization.

208
00:12:49,360 --> 00:12:53,120
So I think that really sort of made it clear that the future was going

209
00:12:53,120 --> 00:12:56,960
to be based on on data. Well, I I like that

210
00:12:56,960 --> 00:13:00,205
you pointed out, the importance of observability.

211
00:13:01,385 --> 00:13:04,765
My career path winding as it was,

212
00:13:05,385 --> 00:13:09,225
I made a a leap from being a software developer to being

213
00:13:09,225 --> 00:13:13,070
a data really a database developer. When I made that

214
00:13:13,070 --> 00:13:16,910
transition, one of the things I had noticed, this was two two and a half

215
00:13:16,910 --> 00:13:20,670
decades ago, I had just started in software development

216
00:13:20,670 --> 00:13:24,510
doing test driven development and it had just

217
00:13:24,510 --> 00:13:28,065
come out, it was called fail first development. I remember thinking

218
00:13:28,065 --> 00:13:31,745
this was perfect. It was a big deal. Yeah. It was. Yeah. Twenty five

219
00:13:31,745 --> 00:13:35,524
years ago. And I remember thinking this is perfect because I'm always failing.

220
00:13:35,985 --> 00:13:39,745
So this this will work nothing ever runs the first time and if it does,

221
00:13:39,745 --> 00:13:43,260
it's suspect. But when I got over into data, I had just

222
00:13:43,260 --> 00:13:46,880
become, you know, kind of a a big believer in the power

223
00:13:47,180 --> 00:13:50,779
and and and really the the confidence that

224
00:13:50,779 --> 00:13:54,240
test driven development gave me. And I was like, we need that

225
00:13:54,700 --> 00:13:58,525
over here. And so it was, just a

226
00:13:58,525 --> 00:14:02,205
field that's fascinating me. I have an engineering background, and so it kind of flowed

227
00:14:02,205 --> 00:14:05,505
right through. Instrumenting the data engineering,

228
00:14:06,525 --> 00:14:10,225
was a big deal so that, again, you could achieve what we now call

229
00:14:10,445 --> 00:14:14,200
observability. But being able to watch that data flow

230
00:14:14,600 --> 00:14:18,440
and when I would mention this to people kinda like you in 2019, I

231
00:14:18,440 --> 00:14:21,800
I would get all sorts of responses. Most of them kinda raised

232
00:14:21,800 --> 00:14:25,640
eyebrows. And I would, some of the more interesting ones

233
00:14:25,640 --> 00:14:29,255
were things along the lines of, well, the data is sort of self

234
00:14:29,255 --> 00:14:32,855
documenting. I mean, it's it's just there. And I'm

235
00:14:32,855 --> 00:14:36,635
like, no. No. It's not. It's I especially when you've moved it through

236
00:14:36,935 --> 00:14:40,775
a bunch of transformation to put it into a business intelligence solution or data

237
00:14:40,775 --> 00:14:44,410
warehouse or or any of that. And that now feeds,

238
00:14:45,029 --> 00:14:48,790
you know, modern LLMs, AI, and and the like, those

239
00:14:48,790 --> 00:14:52,470
same sorts of, I guess, old school processes, I

240
00:14:52,470 --> 00:14:56,310
do. Or at least that's my my understanding. Maybe I'm reading too much into

241
00:14:56,310 --> 00:15:00,055
that, but I love the idea of having observability go

242
00:15:00,055 --> 00:15:03,735
all the way through. You mentioned lineage. That's huge. You wanna make sure that when

243
00:15:03,735 --> 00:15:07,175
you, you know, you make this one change, that's not gonna affect anything

244
00:15:07,175 --> 00:15:10,935
else. Usually, it does affect other things, and having

245
00:15:10,935 --> 00:15:14,750
that lineage view is huge. That is spot on.

246
00:15:14,750 --> 00:15:18,510
That's exactly how we've we've thought about this as well. So, you know, I

247
00:15:18,510 --> 00:15:22,270
think there are specific things that you can test for in data. Like, for

248
00:15:22,270 --> 00:15:25,890
example, you know, specific thing that you can declare, you can say, like,

249
00:15:26,110 --> 00:15:29,755
you know, you know, a T shirt

250
00:15:29,755 --> 00:15:33,514
size should only be, you know, small, medium, large, extra large, whatever.

251
00:15:33,514 --> 00:15:37,274
Right? But then there are some specific things that, you

252
00:15:37,274 --> 00:15:40,654
know, you you don't necessarily know. Like, for example, if there's a particular,

253
00:15:42,000 --> 00:15:45,759
you know, pattern that the data is being updated,

254
00:15:45,759 --> 00:15:49,600
you can actually use machine learning to automatically learn that pattern and then forecast

255
00:15:49,600 --> 00:15:53,440
when it should get up updated again. So it's not necessary for someone to

256
00:15:53,440 --> 00:15:56,420
manually write a test for that. Right? And so

257
00:15:57,015 --> 00:16:00,615
I actually think it's a combination of both of those things which really

258
00:16:00,615 --> 00:16:04,375
give confidence to to data teams over time. So there there's sort of a

259
00:16:04,375 --> 00:16:07,915
couple components to it. The first, I think it really starts with visibility,

260
00:16:08,215 --> 00:16:11,780
sort of call it end to end observability, but it really includes, like, you know,

261
00:16:11,780 --> 00:16:15,620
you mentioned a few of these parts, but, the data

262
00:16:15,620 --> 00:16:18,360
lake, the data warehouse, an orchestration,

263
00:16:19,460 --> 00:16:23,080
BI, ML, AI application that can include the agent,

264
00:16:23,140 --> 00:16:26,925
the vector base if you have a prompt. Right all of those

265
00:16:26,925 --> 00:16:30,764
components you have to have visibility. The first thing is actually to to

266
00:16:30,764 --> 00:16:34,524
your point, like, having lineage into what are the different components that can cross

267
00:16:34,524 --> 00:16:38,045
this. So all the way from. You know, sort of ingestion of the data to

268
00:16:38,045 --> 00:16:41,600
consumption of it. And the second is to start observing.

269
00:16:42,300 --> 00:16:46,140
And and, you know, you there are some specific things that you can declare

270
00:16:46,140 --> 00:16:49,580
and test and based on your business needs, and there are some things that you

271
00:16:49,580 --> 00:16:52,700
can do in an automated way. And and, actually, I think this is an area

272
00:16:52,700 --> 00:16:55,595
where AI can help. So for example,

273
00:16:56,695 --> 00:17:00,375
what what oftentimes teams end up doing is spending a lot of time

274
00:17:00,375 --> 00:17:03,975
trying to define what are data quality rules. And,

275
00:17:03,975 --> 00:17:07,115
actually, you can use LLMs to profile the data,

276
00:17:07,800 --> 00:17:10,619
Make some make some, yeah, make some inference,

277
00:17:11,560 --> 00:17:14,700
based on the semantic meaning of data and then make recommendations.

278
00:17:15,560 --> 00:17:19,240
So for example, I I love this example. We work with lots

279
00:17:19,240 --> 00:17:23,015
of, sports teams. And so you can imagine that,

280
00:17:23,015 --> 00:17:26,775
you know, you have a particular field called, like, let's say this is

281
00:17:26,775 --> 00:17:30,235
in baseball, a baseball team and sort of, like, you know, pitch type.

282
00:17:30,615 --> 00:17:34,375
And and then, like, the the speed that matches that. And

283
00:17:34,375 --> 00:17:38,070
so you can imagine that, like, an l m can recommend or infer that

284
00:17:38,070 --> 00:17:41,750
a fastball should not be, you know, less than

285
00:17:41,750 --> 00:17:45,510
70 miles per hour or whatever it is. Even though I don't know what

286
00:17:45,510 --> 00:17:48,845
the real number is. I just made that up. But there is, like, some you

287
00:17:48,845 --> 00:17:52,605
you can infer based based on that and make a recommendation. And

288
00:17:52,605 --> 00:17:56,045
so, actually, it's a I find that AI and LM is a really cool

289
00:17:56,045 --> 00:17:59,885
application of how to make observability faster and and and

290
00:17:59,885 --> 00:18:03,590
easier for for teams. So, yeah, I'm I'm

291
00:18:03,590 --> 00:18:07,290
very excited about about what you just shared, Andy. Well,

292
00:18:07,430 --> 00:18:11,110
I I love what you brought up about machine learning being able to to

293
00:18:11,110 --> 00:18:14,870
make basically make predictions about things.

294
00:18:14,870 --> 00:18:18,090
And and one of the terms that, you know, as a practitioner

295
00:18:18,765 --> 00:18:22,524
of, business intelligence is especially the data engineering that supports

296
00:18:22,524 --> 00:18:26,285
it Mhmm. Is data volatility. Mhmm. So if I'm

297
00:18:26,525 --> 00:18:29,885
especially if I'm looking at an outlier. So I'm consuming this

298
00:18:29,885 --> 00:18:33,730
data day in and day out, And let's

299
00:18:33,730 --> 00:18:37,350
say, you know, 10% of the data is new stuff,

300
00:18:37,730 --> 00:18:41,410
and maybe another 10 or 15% are things that are have

301
00:18:41,410 --> 00:18:45,030
been updated, old stuff that's been updated, and the rest of it's relatively

302
00:18:45,090 --> 00:18:48,735
stable. If I see those numbers go crazy out of bounds,

303
00:18:49,195 --> 00:18:52,795
you know, and machine learning would be able to pick that up right

304
00:18:52,795 --> 00:18:56,555
away and say, there may be a problem with the data we're

305
00:18:56,555 --> 00:19:00,330
reading today. You know, I would I that that sounds like one of

306
00:19:00,330 --> 00:19:02,509
the problems that would solve is that volatility,

307
00:19:03,769 --> 00:19:07,370
expected ranges of volatility of data. That's exactly

308
00:19:07,370 --> 00:19:11,095
right. Yeah. Cool. Interesting. I think there's

309
00:19:11,095 --> 00:19:14,934
also something you said was, you know, when you have LLMs, because, obviously, we have

310
00:19:14,934 --> 00:19:18,774
to talk about GenAI because it's 2025, and I think you're in

311
00:19:18,774 --> 00:19:22,534
Silicon Valley. I think if you don't mention GenAI every twenty five

312
00:19:22,534 --> 00:19:25,809
minutes, the cops come and knock on your door and check it out. Welfare check.

313
00:19:25,890 --> 00:19:29,650
Could get in trouble. Or they make sure you're okay. Make

314
00:19:29,650 --> 00:19:33,490
sure you're okay. But I think one of the things that really

315
00:19:33,490 --> 00:19:36,929
kind of makes me worry about GenAI is that it's not

316
00:19:36,929 --> 00:19:40,065
immediately obvious. Like, if you're at the airport, obviously, it's not a good look for

317
00:19:40,065 --> 00:19:42,865
you. Like, if the if the and this has happened to me where the app

318
00:19:42,865 --> 00:19:46,545
says one thing, the screen says something else, and my ticket says yet a

319
00:19:46,545 --> 00:19:50,005
third thing. So I'm not really sure where I'm supposed to go.

320
00:19:50,145 --> 00:19:53,205
Generally speaking of those, the app tends to be more accurate.

321
00:19:53,880 --> 00:19:57,100
But, that depends on the airline.

322
00:19:57,400 --> 00:20:00,860
But with with LLMs, it's a the latency

323
00:20:00,920 --> 00:20:04,440
between you seeing the data where the cons the bad

324
00:20:04,440 --> 00:20:07,320
consequences of the data tends to be a lot more

325
00:20:08,674 --> 00:20:12,355
I'll use a $10 word today. I can't even say

326
00:20:12,355 --> 00:20:16,034
it, but it's not it's not immediately obvious. Right? There goes my

327
00:20:16,034 --> 00:20:19,315
my fail and my $10 word. But, like, it's not like it there's a lot

328
00:20:19,315 --> 00:20:23,070
more steps in labyrinthine. I'll go with that one because I can say that.

329
00:20:23,070 --> 00:20:26,289
But, like, what so how do you provide

330
00:20:26,590 --> 00:20:29,330
observability in something like LLMs where

331
00:20:30,509 --> 00:20:34,350
the, the input and the output time tends to not

332
00:20:34,350 --> 00:20:37,970
be quite as straightforward as a data as an old school data pipeline?

333
00:20:38,625 --> 00:20:42,085
Yeah. Such a great question. And maybe I'll just share some of my favorite

334
00:20:43,025 --> 00:20:46,385
wonders if that's helpful. And and I think I'll share them

335
00:20:46,385 --> 00:20:50,225
because it's helpful to explain the gravity

336
00:20:50,225 --> 00:20:53,185
of these issues. So, for example, you know, if you're in an airport and, you

337
00:20:53,185 --> 00:20:56,560
know, the app doesn't say the same as what you have,

338
00:20:56,860 --> 00:20:59,980
hopefully, you arrive early at airports, Frank. I don't know if you have enough time

339
00:20:59,980 --> 00:21:03,600
to, like, figure out the discrepancy and you won't miss your flight. Right?

340
00:21:04,700 --> 00:21:07,760
But oftentimes, those things can lead to to really big disasters.

341
00:21:08,380 --> 00:21:11,914
Even three gen AI. So so I think this was in 2020.

342
00:21:12,294 --> 00:21:15,895
Unity, which is a gaming company, they had one schema

343
00:21:15,895 --> 00:21:19,515
change, resulting in a hundred million dollar loss.

344
00:21:19,655 --> 00:21:23,255
Their stock dropped 37%. Oh my gosh. Pretty

345
00:21:23,255 --> 00:21:26,990
meaningful. Right? Fast forward, I think this was

346
00:21:26,990 --> 00:21:30,770
2023 or 2024,

347
00:21:31,470 --> 00:21:33,810
but not so much related to AI yet.

348
00:21:34,830 --> 00:21:38,415
Citibank was hit with a $400,000,000 fine for

349
00:21:38,535 --> 00:21:42,095
I remember that. For data quality practices for lack

350
00:21:42,095 --> 00:21:45,275
of data quality practices. So think about all the regulatory

351
00:21:46,135 --> 00:21:49,835
industries like health care, financial services,

352
00:21:50,055 --> 00:21:53,840
like, you know, wherever there's, like, PII and and,

353
00:21:55,760 --> 00:21:59,360
And and the, like, you know, the the

354
00:21:59,360 --> 00:22:03,140
implication there are pretty grave. Some fun examples for more recently.

355
00:22:03,200 --> 00:22:06,785
I don't know if fun. I shouldn't call them fun. Some other examples from

356
00:22:06,945 --> 00:22:10,625
yeah. You mentioned Chevy. So I think there was a user

357
00:22:10,625 --> 00:22:14,325
that convinced a chatbot to sell the Chevy Tahoe

358
00:22:14,545 --> 00:22:18,385
for $1. I I commend the user from being able to

359
00:22:18,385 --> 00:22:22,090
do that, but that is terrible. Right? That's terrible

360
00:22:22,090 --> 00:22:25,550
that, that happened. And that chatbot went down

361
00:22:25,930 --> 00:22:28,890
the next day. They they took it offline the next day. I think it was

362
00:22:28,890 --> 00:22:31,630
in Fremont, California, so not that far from the bay.

363
00:22:32,505 --> 00:22:35,725
Yeah. So right. So that's pretty pretty consequential.

364
00:22:37,385 --> 00:22:41,185
I'll just give another, like, example. This is my favorite example. This is what

365
00:22:41,305 --> 00:22:45,145
it went viral on x couple months ago. Someone googled, what should I

366
00:22:45,145 --> 00:22:48,900
do when cheese is slipping off my pizza? And Google responded,

367
00:22:48,960 --> 00:22:50,740
oh, you should just use organic superglue.

368
00:22:54,080 --> 00:22:57,840
Great answer. They they had some really good gaps.

369
00:22:57,840 --> 00:23:01,200
There was the, eat eat one rock a day to get your,

370
00:23:01,680 --> 00:23:05,385
minerals and stuff like that. Yeah. So I I

371
00:23:05,385 --> 00:23:09,145
love that because that's an example of where, like, the prompt was

372
00:23:09,145 --> 00:23:12,745
fine, the context was probably fine, the model was

373
00:23:12,745 --> 00:23:16,284
fine, but the model output was totally not fine.

374
00:23:16,620 --> 00:23:20,460
Right? Right. And so and by the way, maybe Google can get away with it

375
00:23:20,460 --> 00:23:24,140
because it's Google, but, like, 99.9% of brands can't get

376
00:23:24,140 --> 00:23:27,980
away with with the mistakes. Right? And so what, you know, what

377
00:23:27,980 --> 00:23:31,100
do you do? How do you provide observability in in that world? What does that

378
00:23:31,100 --> 00:23:34,875
look like? First, I'll just say, I think

379
00:23:34,875 --> 00:23:38,575
there's still human in the loop, and there will be. So, actually, you know,

380
00:23:38,635 --> 00:23:42,395
it's interesting going back to 2019 when we started the company. People would tell us,

381
00:23:42,395 --> 00:23:45,930
oh, you know, I have this important report that my CEO looks at.

382
00:23:46,090 --> 00:23:49,770
But before they look at it, I have, like, six different people looking at the

383
00:23:49,770 --> 00:23:52,890
report with, like, you know, sets of eyes to make sure that the data is

384
00:23:52,890 --> 00:23:56,730
accurate. So, like, people use manual stuff back then. Today, what I

385
00:23:56,730 --> 00:24:00,575
hear is I was just speaking with this head of AI, Silicon Valley,

386
00:24:00,875 --> 00:24:03,595
and I was like, how do you make sure the answers are accurate? And they

387
00:24:03,595 --> 00:24:07,275
were like, well, we have someone sifting through dozens, hundreds of

388
00:24:07,275 --> 00:24:10,955
responses every single day to make sure they're accurate. So I don't think human in

389
00:24:10,955 --> 00:24:14,475
the loop evaluation is going anywhere. There's more advanced techniques, you know,

390
00:24:14,475 --> 00:24:18,310
comparing to to to ground truth data, using LLM

391
00:24:18,310 --> 00:24:22,150
as a judge. There's various sort of, things that we can do, but but I

392
00:24:22,150 --> 00:24:25,850
think human isn't going away. In terms of observability,

393
00:24:27,270 --> 00:24:30,855
I talked before I'll explain a little bit about this sort of framework

394
00:24:30,914 --> 00:24:34,434
of, you know, data issues can be really traced back

395
00:24:34,434 --> 00:24:37,955
to these four core root causes, and I think it's

396
00:24:37,955 --> 00:24:41,635
important to have observability for each in in sort of this world.

397
00:24:41,635 --> 00:24:45,414
So the first I mentioned is data. And so by that, I mean,

398
00:24:45,970 --> 00:24:49,810
you know, let's use another example. Credit Karma, for example,

399
00:24:49,810 --> 00:24:53,590
has a financial advisor chatbot where, basically, they take in information

400
00:24:53,650 --> 00:24:57,330
about you that they have, you know, like, what kind of car you

401
00:24:57,330 --> 00:25:00,370
have as being of cars and, you know, where you live and whatnot, and then

402
00:25:00,370 --> 00:25:04,125
they make financial recommendations based on that. If the

403
00:25:04,125 --> 00:25:07,725
data that they are ingesting from third party data is late or isn't

404
00:25:07,725 --> 00:25:11,485
arriving or is incomplete, that messes up everything downstream. So one

405
00:25:11,485 --> 00:25:14,685
root cause can be the data that you're ingesting is just wrong. Maybe it's all

406
00:25:14,685 --> 00:25:18,309
null values, for example. The second can

407
00:25:18,309 --> 00:25:22,070
be due to change in the code. So the code could be like a a

408
00:25:22,070 --> 00:25:25,830
bad like a schema change, like in the Unity example. It could be a change

409
00:25:25,830 --> 00:25:29,190
in the code that's actually, being used for the

410
00:25:29,190 --> 00:25:32,865
agent. Really, code change can happen every anywhere. And, by the

411
00:25:32,865 --> 00:25:36,385
way, not necessarily by the data team. It can happen by an engineering team or

412
00:25:36,385 --> 00:25:40,225
someone else. It has nothing to do with the with the data state. Right? So

413
00:25:40,225 --> 00:25:43,820
code changes can contribute. The third is system.

414
00:25:44,280 --> 00:25:48,040
A % of systems fail. What what do I mean by system? I

415
00:25:48,040 --> 00:25:51,580
mean system is, like, basically the infrastructure that sort of runs all these jobs.

416
00:25:51,880 --> 00:25:55,559
So this could be, like, an airflow job that fails or a DDT job

417
00:25:55,559 --> 00:25:59,255
that that fails. You know, again, a % of systems fail,

418
00:25:59,255 --> 00:26:02,795
and so you would definitely have something that goes wrong in systems.

419
00:26:03,095 --> 00:26:06,295
And then the fourth is you could just have the model output be wrong, kinda

420
00:26:06,295 --> 00:26:10,135
like with the cheese in in Google, example. And

421
00:26:10,135 --> 00:26:13,919
so when we think about sort of having what does it mean,

422
00:26:13,919 --> 00:26:17,760
what does observability mean in this in this age, I think it has to

423
00:26:17,760 --> 00:26:21,380
have coverage for all four of those things. And here's the problem. It oftentimes

424
00:26:21,600 --> 00:26:25,280
includes all four together. So I don't know if it you know, it's typically on

425
00:26:25,280 --> 00:26:28,945
a Friday at 5PM. You're just about done, and then

426
00:26:29,164 --> 00:26:33,005
everything breaks at the same time. That's an

427
00:26:33,005 --> 00:26:36,765
interesting point. Like and and it's you also use the a term

428
00:26:36,765 --> 00:26:40,605
a couple of times, which, you're I can count on one hand how many

429
00:26:40,605 --> 00:26:42,950
non Microsoft people have used this term,

430
00:26:44,450 --> 00:26:47,970
data estate. And I'm just curious about I know where I pick from

431
00:26:47,970 --> 00:26:51,809
Microsoft. No. No. No. Like, I'm like I mean, I always

432
00:26:51,809 --> 00:26:55,405
thought it was a, you know, Microsoft invention. I don't think it is.

433
00:26:55,465 --> 00:26:59,145
But, like, where did you pick up that term? Because I've only like, seriously, you

434
00:26:59,145 --> 00:27:02,825
were, like, the third or maybe fourth person who is not

435
00:27:02,985 --> 00:27:06,105
never worked for Microsoft, never worked with Microsoft. I I mean, I don't know if

436
00:27:06,105 --> 00:27:09,909
you work with Microsoft, but, like, I I always whenever I hear someone say

437
00:27:09,909 --> 00:27:13,029
data to state publicly, I'm like, so who'd you work for at Microsoft? What division?

438
00:27:13,029 --> 00:27:16,789
Like, like Oh, wow. Yeah. It's like that. And at first, I

439
00:27:16,789 --> 00:27:19,830
didn't like I'll be honest. I didn't like the term at all, but eventually, I

440
00:27:19,830 --> 00:27:23,405
kinda grew to like the term because it there's a lot behind it, and I'd

441
00:27:23,405 --> 00:27:27,245
be curious to get, like, one, where'd you where'd you where'd you

442
00:27:27,245 --> 00:27:30,845
pick that up? Like, I'm just, like and then two, what does it mean to

443
00:27:30,845 --> 00:27:34,605
you? Like, what does that term data state mean to you? Great question. For

444
00:27:34,605 --> 00:27:38,320
what it's worth, I actually didn't like it either. For the record, I didn't even

445
00:27:38,320 --> 00:27:41,840
like data observability to begin with Mhmm. To be totally Really? English is

446
00:27:42,000 --> 00:27:45,360
yeah. English is my second language, and observability was such a difficult word to

447
00:27:45,360 --> 00:27:49,205
pronounce. When we started the when we started the, you know,

448
00:27:49,205 --> 00:27:51,845
the company and and the category, we had to give it a name. So we

449
00:27:51,845 --> 00:27:55,365
didn't really know is this you know, we used we we coined the term data

450
00:27:55,365 --> 00:27:59,045
downtime, you know, as a corollary to application downtime. We thought maybe

451
00:27:59,045 --> 00:28:02,565
data reliability. There are lots of

452
00:28:02,565 --> 00:28:06,320
options. At the end of the day, I always try to get gravitate towards where

453
00:28:06,320 --> 00:28:10,080
my customers are, so whatever language my customers use. And so customers

454
00:28:10,080 --> 00:28:13,440
started using the word observability, so I started using that too. And same with the

455
00:28:13,440 --> 00:28:17,039
state, they started using the data state sort of as a language. And so

456
00:28:17,360 --> 00:28:21,065
Interesting. Full disclosure, have not, have no

457
00:28:21,065 --> 00:28:24,684
ties to Microsoft, but but just have heard

458
00:28:24,825 --> 00:28:28,365
mostly enterprises sort of think about that. I I think my understanding,

459
00:28:28,745 --> 00:28:32,470
you know, for for what they mean is, you know, wherever

460
00:28:32,470 --> 00:28:35,830
you store aggregate process data. And so that, you know, can

461
00:28:35,830 --> 00:28:39,450
include, you know, you know, upstream

462
00:28:39,590 --> 00:28:43,269
sources or upstream, data sources. But, you know, it could be,

463
00:28:43,269 --> 00:28:46,885
like, an Oracle or SAP database. It could be data

464
00:28:46,885 --> 00:28:49,865
lake house, data warehouse like Snowflake, Databricks,

465
00:28:51,205 --> 00:28:54,885
AWS, Redshift, s three, all the

466
00:28:54,885 --> 00:28:58,405
way to wherever you're consuming that. That could be a BI report. You know, Power

467
00:28:58,405 --> 00:28:59,760
BI. Sorry, Microsoft.

468
00:29:02,140 --> 00:29:05,679
Right, Looker, Tableau, you know,

469
00:29:06,460 --> 00:29:09,980
various, various options. And,

470
00:29:09,980 --> 00:29:13,740
honestly, the, you know, the most common enterprise has all of

471
00:29:13,740 --> 00:29:17,434
the above in some shape or forward fashion. And so to sort

472
00:29:17,434 --> 00:29:21,195
of include all of that, I think

473
00:29:21,195 --> 00:29:25,035
the some of the thesis that we have around observability is that, by the way,

474
00:29:25,035 --> 00:29:28,815
each of those by themselves has some concept of observability.

475
00:29:29,115 --> 00:29:32,680
Right? Like, you

476
00:29:32,680 --> 00:29:35,900
can, for example, with Snowflake, you can set up some basic,

477
00:29:36,600 --> 00:29:40,280
sort of checks, if you will, like a sum check or whatever. Right?

478
00:29:40,280 --> 00:29:43,855
You you could do that in Snowflake. However, we think that observability

479
00:29:43,995 --> 00:29:47,674
needs to be sort of third party and to be end to end. And,

480
00:29:47,674 --> 00:29:51,375
again, that draws on on software corollary. So,

481
00:29:51,595 --> 00:29:55,034
you know, like, AWS has CloudWatch, for example,

482
00:29:55,034 --> 00:29:58,830
but that's probably not sufficient for whatever you're building. You're probably

483
00:29:58,830 --> 00:30:02,210
gonna use, again, like, New Relic or Datadog to connect

484
00:30:02,590 --> 00:30:05,650
across the the board to, you know, variety of of,

485
00:30:06,830 --> 00:30:10,644
integrations. Right? They have hundreds. So that's what I think about when I

486
00:30:10,644 --> 00:30:14,245
say data estate. But it's a great question. It's definitely not my

487
00:30:14,245 --> 00:30:17,524
word. No. I was just curious. Like like, you know,

488
00:30:17,924 --> 00:30:21,445
because whenever because first, I hated the term too. Right? And I can't maybe it's

489
00:30:21,445 --> 00:30:23,420
Stockholm Syndrome. I don't know. But,

490
00:30:26,040 --> 00:30:29,400
the more I kind of sat on it and kind of digested it, I was

491
00:30:29,400 --> 00:30:32,780
like, I like it because it explains, like, you know, you know, historically.

492
00:30:33,000 --> 00:30:36,760
Right? Like, a state is, you know, whoever

493
00:30:36,760 --> 00:30:39,674
owned the land got to call the shots and whoever called the shots owned the

494
00:30:39,674 --> 00:30:42,795
land. Like, there was a very, you know, you drew the food, you you cut

495
00:30:42,795 --> 00:30:46,635
down the trees, you, you know, you mined for, I think the Minecraft

496
00:30:46,635 --> 00:30:50,155
movie is coming out. So you mined for all these things. Right? My kids are

497
00:30:50,155 --> 00:30:53,980
into it. But, like, and it's

498
00:30:53,980 --> 00:30:57,179
really kinda like it's just the idea of seeing it, like, it's land. It's kinda

499
00:30:57,179 --> 00:31:00,860
like land. It's kinda like a natural resource. It's not really natural, but it is

500
00:31:00,860 --> 00:31:04,620
a resource. Right? And if I say unnatural resource, that's really weird. But it's a

501
00:31:04,620 --> 00:31:08,140
resource. Right? And if you you can either you have it. You already have

502
00:31:08,140 --> 00:31:11,945
it. You either develop it or you don't. And, you know, do

503
00:31:11,945 --> 00:31:15,465
you, you know, do you grow food on it? Do you, you know, like so

504
00:31:15,465 --> 00:31:19,305
see, I I liked it because it was the idea that it's already there. Right?

505
00:31:19,305 --> 00:31:22,425
Mhmm. And it's it might be in forms you don't really think about. Right? Like,

506
00:31:22,505 --> 00:31:26,080
you know, PDFs in a in a SMB share somewhere.

507
00:31:26,080 --> 00:31:29,460
Right? Mhmm. I mean, that's part of your data to state. Yep. Right?

508
00:31:30,559 --> 00:31:34,240
And it's that's how I kinda, like, came to terms with it. And,

509
00:31:34,240 --> 00:31:37,520
like, I really kinda like it because it helps you to think holistically about data

510
00:31:37,520 --> 00:31:40,934
because I think a lot of business decision

511
00:31:40,934 --> 00:31:44,775
makers and even technical decision makers don't see data as a

512
00:31:44,775 --> 00:31:47,995
as a as a as a resource. I think that's changed

513
00:31:48,775 --> 00:31:51,434
over the last maybe five, six years.

514
00:31:52,934 --> 00:31:56,730
But it really became something that they don't see

515
00:31:56,730 --> 00:31:59,929
it as a resource they could mine, they can get value out of. Right? The

516
00:31:59,929 --> 00:32:03,530
smart people did. But, for the most part That's

517
00:32:03,530 --> 00:32:07,309
right. Yeah. You had to convince them. Right? Exactly.

518
00:32:07,450 --> 00:32:10,784
It sounds like based on what you say because, like, you know, my wife works

519
00:32:10,784 --> 00:32:14,225
in IT security. Right? So, so we're a two engineer

520
00:32:14,225 --> 00:32:18,065
household. So the kids are super nerds. But, like, I was telling

521
00:32:18,065 --> 00:32:21,505
her after chat CPT came out, I was all excited about it. And I was

522
00:32:21,505 --> 00:32:24,304
telling her about how this works. I was like, you give it this big corpus

523
00:32:24,304 --> 00:32:27,330
of data, and they chews through it, and it comes up with these these vectors

524
00:32:27,330 --> 00:32:30,450
and stuff like that. And then she looked at me and it's like, so all

525
00:32:30,450 --> 00:32:33,030
the training data is now a massive attack surface.

526
00:32:34,450 --> 00:32:38,210
And Yep. When that's just why I love my wife. So I

527
00:32:38,450 --> 00:32:42,245
I'm wronged. She's never wronged. Well, that's true. But at

528
00:32:42,245 --> 00:32:45,765
first I was like I was thinking but but you're missing and then I was

529
00:32:45,765 --> 00:32:47,925
gonna say you're missing the point which one is never a good thing to say

530
00:32:47,925 --> 00:32:51,684
but Like midway through I was like, oh my gosh,

531
00:32:51,684 --> 00:32:54,640
she's right. Oh my gosh. She's right. So then

532
00:32:55,660 --> 00:32:59,020
when I started talking to other data science and AI types, and I was like,

533
00:32:59,020 --> 00:33:02,540
but but don't you think this could be, like, a big attack surface? I look

534
00:33:02,540 --> 00:33:06,220
like that meme with the guy from It's Sunny in Philadelphia with, like, it's

535
00:33:06,220 --> 00:33:10,005
always sunny where he had, like, the conspiracy thing. Like, I swear I will

536
00:33:10,005 --> 00:33:13,765
like that meme. Yeah. And, you know, and if you

537
00:33:13,765 --> 00:33:17,545
look at the I think OWASP has, like, the top 10 vulnerabilities of LLMs

538
00:33:17,605 --> 00:33:21,190
that is either two or three. Right? So it's

539
00:33:21,190 --> 00:33:22,970
kinda like there's a fine line between,

540
00:33:24,870 --> 00:33:28,549
like, thinking too much about problem, but also kind of thinking ahead of the

541
00:33:28,549 --> 00:33:32,309
problem. I don't know. No. Oh, I think you

542
00:33:32,309 --> 00:33:35,135
cut off a little bit, Frank, but, Andy,

543
00:33:36,315 --> 00:33:39,355
to me, that resonates a lot, and I think it's sort of really the overlap

544
00:33:39,355 --> 00:33:43,035
between data and engineers. And, by the way, like, we didn't even talk

545
00:33:43,035 --> 00:33:46,635
about security. Like, all these concepts also exist in security.

546
00:33:46,635 --> 00:33:49,870
Right? And I think in the same way that we sort of manage, like, you

547
00:33:49,870 --> 00:33:53,470
know, sub zero, sub one issues in security engineering, data

548
00:33:53,470 --> 00:33:56,429
issues should be treated the same way. You should have a framework to understand what's

549
00:33:56,429 --> 00:33:59,789
a sub zero, what's a sub one for data issues. You should it should be

550
00:33:59,789 --> 00:34:02,669
connected to pager duty. Like, people should wake up in the middle of the night

551
00:34:02,669 --> 00:34:06,434
when you have data issues. I think I think that's right. It's

552
00:34:06,434 --> 00:34:10,034
improving, but, we're not quite there. It'll

553
00:34:10,034 --> 00:34:13,635
happen. No. You're right, though. Like, they don't think about this in

554
00:34:13,635 --> 00:34:17,475
terms of they don't does it I wouldn't say it's not disciplined. Sorry,

555
00:34:17,475 --> 00:34:20,909
Annie. I cut you off. No. But my experience we talked to data engineers. Sorry,

556
00:34:20,909 --> 00:34:24,510
Andy. And I I I I am a former data engineer

557
00:34:24,510 --> 00:34:28,049
myself. Like, I thought of it in terms of schema structures and pipelines.

558
00:34:28,270 --> 00:34:32,109
Mhmm. Not necessarily securing those pipelines. Right? Mhmm. Sorry,

559
00:34:32,109 --> 00:34:35,885
Andy. I'll go. No. I was curious. I wanted to to shift back

560
00:34:35,885 --> 00:34:39,344
to you. You mentioned the four areas that your software,

561
00:34:39,885 --> 00:34:43,485
looks over your AI and the observability software does. What

562
00:34:43,485 --> 00:34:45,985
happens when it detects something amiss?

563
00:34:47,589 --> 00:34:51,369
Great question. So not even talking about Monte Carlo specifically, but rather

564
00:34:51,510 --> 00:34:55,349
an observability solution. I think an observability solution needs to

565
00:34:55,349 --> 00:34:59,109
have coverage or an observability approach, by the way. Like, some people build this

566
00:34:59,109 --> 00:35:02,715
in house. An observability approach should take into consideration

567
00:35:03,415 --> 00:35:07,255
your data estate, should take into consideration, right, your

568
00:35:07,255 --> 00:35:11,015
entire data estate. I think, oftentimes, the mistake is people will even if they

569
00:35:11,015 --> 00:35:13,975
build it in house or do anything else, they'll really just focus on, like, the

570
00:35:13,975 --> 00:35:17,819
data and their data lake or the data in a particular report. Like, that's

571
00:35:17,819 --> 00:35:21,660
not sufficient. Right? It it just isn't. And so people waste

572
00:35:21,660 --> 00:35:24,940
a ton of time trying to understand, like, what's wrong and where. So I think

573
00:35:24,940 --> 00:35:28,700
the first is, like, you need you need visibility across the data

574
00:35:28,700 --> 00:35:32,355
state, which hopefully we've defined an unnatural resource that should be

575
00:35:32,355 --> 00:35:36,115
managed securely. And and I think that's right because I

576
00:35:36,115 --> 00:35:39,655
I by the way, Monte Carlo doesn't doesn't do the security

577
00:35:39,715 --> 00:35:43,415
part, but I similarly believe that in the same kind of diligence

578
00:35:43,475 --> 00:35:47,080
that we apply to data as engineering, you want data products to

579
00:35:47,080 --> 00:35:50,619
be reliable but also secure, scalable,

580
00:35:50,840 --> 00:35:54,520
like all those concepts should adapt. By chance, we happen to

581
00:35:54,520 --> 00:35:57,720
focus on the reliability and observability part, but all the other,

582
00:35:58,615 --> 00:36:00,955
principles of software engineering should apply.

583
00:36:02,295 --> 00:36:06,135
We specifically don't do it, but very much believe that should be

584
00:36:06,135 --> 00:36:09,895
the case. But back to your question, you

585
00:36:09,895 --> 00:36:13,195
know, so so what happens when there is an issue?

586
00:36:13,970 --> 00:36:17,730
Very similar to workflow that you might find in Datadog,

587
00:36:17,730 --> 00:36:21,410
New Relic, and and PagerDuty. So there is an alert that goes out,

588
00:36:21,730 --> 00:36:25,329
often you know, in whatever flavor of choice. If you're an enterprise that has a

589
00:36:25,329 --> 00:36:28,975
data state, this is likely Microsoft Teams. If not, this would mean

590
00:36:28,975 --> 00:36:32,735
Slack or an email or what you know, some teams like to have it connected

591
00:36:32,735 --> 00:36:36,335
to to Jira and and pager duty for for sev zeros or sev

592
00:36:36,335 --> 00:36:40,175
ones. And, you know, the first thing

593
00:36:40,175 --> 00:36:43,810
that people will do is start, you know, typically an analyst.

594
00:36:43,870 --> 00:36:47,070
I was I was in, you know, prior an analyst. The first thing you start

595
00:36:47,070 --> 00:36:49,970
asking yourself is, why the hell is the data is wrong?

596
00:36:50,590 --> 00:36:53,970
Right. Yeah. You're like, well, was the report on time?

597
00:36:54,234 --> 00:36:58,075
Was the data accurate? Was it complete? You start going through all

598
00:36:58,155 --> 00:37:01,915
and then you start you basically come up with hypothesis. And then you start

599
00:37:01,915 --> 00:37:05,755
researching those hypothesis, and you're like, well, let me let me

600
00:37:05,755 --> 00:37:09,070
trace the data all the way all the steps of the transformation

601
00:37:09,450 --> 00:37:12,570
and start looking. Was the data okay here? Yes. Check. Okay. Move on. Was it

602
00:37:12,570 --> 00:37:16,110
data right? You literally you started this, like, recursive process. Gotcha.

603
00:37:16,490 --> 00:37:20,010
Before we started the company, I used to do this all manually. So I remember,

604
00:37:20,010 --> 00:37:22,895
like, I would go into a, you know, into a room. Maybe you did this

605
00:37:22,895 --> 00:37:26,655
too. And, like, on a whiteboard, I would start, like, basically mapping out

606
00:37:26,655 --> 00:37:30,214
the lineage. Okay. This broke here. Was the data here okay? Let's let

607
00:37:30,335 --> 00:37:33,615
let's sample the data and make sure it's okay. Okay. Move on. Let's like, literally,

608
00:37:33,615 --> 00:37:37,070
we have this, like, very every morning, actually, you know, that this

609
00:37:37,070 --> 00:37:40,670
became such such a problem because we were so reliant on this particular day

610
00:37:40,750 --> 00:37:44,430
dataset that every morning, me and my team would wake up, and we would basically

611
00:37:44,430 --> 00:37:47,890
go step by step and diligently, like, make sure that the data is accurate,

612
00:37:48,110 --> 00:37:51,785
which I felt like was I was like, this is, like, total, you know, crazy.

613
00:37:52,085 --> 00:37:55,684
So, you know, I think, particularly in Monte

614
00:37:55,684 --> 00:37:58,744
Carlo or, like, what observability does is provides the

615
00:37:59,204 --> 00:38:02,805
information that you need in order to troubleshoot and understand where the issue is. And

616
00:38:02,805 --> 00:38:06,599
so we can surface you information like, hey. There was at the same time that

617
00:38:06,599 --> 00:38:10,119
this dataset you know, maybe the the percentage of null values in

618
00:38:10,119 --> 00:38:13,480
particular field was inaccurate. And then at the same time, there was a full

619
00:38:13,480 --> 00:38:17,099
request that happened. Maybe those are correlated, actually. Gotcha.

620
00:38:17,319 --> 00:38:21,135
Maybe, you know and maybe, actually, you can use you can also

621
00:38:21,135 --> 00:38:24,895
do a code analysis. So you can, like, basically, you know, analyst

622
00:38:24,895 --> 00:38:27,615
what we used to do is, like, sift through lines of code and try to

623
00:38:27,615 --> 00:38:30,335
see what the change. Hey. Why did few surface to you that, like, there was

624
00:38:30,335 --> 00:38:33,395
a particular change in the, you know, name of a field,

625
00:38:34,290 --> 00:38:38,050
at the same time as an example. So bringing all that data into one

626
00:38:38,050 --> 00:38:41,350
place can help you sort of troubleshoot that. And

627
00:38:42,050 --> 00:38:45,570
sorry for another LLM plug, but you can actually have

628
00:38:45,570 --> 00:38:49,110
an LLM do this for you, which is pretty sick where it's like an early

629
00:38:49,415 --> 00:38:53,095
beta test for us. We haven't released it yet. But, basically, what we're

630
00:38:53,095 --> 00:38:56,795
testing internally is for every like, for data incidents,

631
00:38:57,255 --> 00:39:00,855
there's basically, like, an in like, a troubleshooting agent that

632
00:39:00,855 --> 00:39:04,503
spawns agents for each of the hypothesis. So there's, like, an agent that

633
00:39:04,770 --> 00:39:08,370
statement. Yeah. I it's really cool. There's an agent that

634
00:39:08,370 --> 00:39:11,890
looks into, like, the code change, the data change, the system

635
00:39:11,890 --> 00:39:15,570
change, and then and then it does it recursively on

636
00:39:15,570 --> 00:39:18,930
all those tables. So you can actually run up to a hundred agents in under

637
00:39:18,930 --> 00:39:22,744
one minute. And then there's a larger LLM that takes all that information

638
00:39:22,744 --> 00:39:26,585
and summarizes it and synthesizes it. So, again, early days, this is like we're still

639
00:39:26,585 --> 00:39:30,345
building it. Very cool. But the early results are really cool. Yeah. It's

640
00:39:30,345 --> 00:39:33,785
like basically turbocharging your your data analysts and your data

641
00:39:33,785 --> 00:39:37,360
stewards. Sorry. I got all excited. No. It's it is That's really

642
00:39:37,360 --> 00:39:40,880
cool. Fascinating, and I love that you're excited about it. And what one of the

643
00:39:40,880 --> 00:39:44,720
jokes that I make when I'm I'm working with my kids on something, if

644
00:39:44,720 --> 00:39:48,275
they nail something, I'll I'll say to them, you know,

645
00:39:48,275 --> 00:39:52,115
something similar to this. It's like, if you can only, you know, if you

646
00:39:52,115 --> 00:39:55,634
can only run a hundred in one minute, I guess that's if that's the best

647
00:39:55,634 --> 00:39:59,095
you can do, we'll just have to live with it. Yeah. Exactly.

648
00:40:00,355 --> 00:40:04,190
That's that's an amazing stat. Yeah. Yeah. That is interesting. And I

649
00:40:04,190 --> 00:40:08,030
also think too I also think too that, like, observability could help

650
00:40:08,030 --> 00:40:11,550
with secure the security story. Right? Because if, you know, you're looking at a

651
00:40:11,550 --> 00:40:15,309
pipeline and it's like, hey. Weren't there a bunch of

652
00:40:15,309 --> 00:40:18,935
sketchy looking IPs, like, poking around our system about the time that this

653
00:40:18,935 --> 00:40:22,695
pipeline ran? Maybe the rest of the data that goes out of that pipeline

654
00:40:22,695 --> 00:40:26,215
run is a little bit suspicious too. Yeah. A

655
00:40:26,215 --> 00:40:29,575
%. Like, we we you know, for example, you work with a,

656
00:40:30,695 --> 00:40:34,360
call it delivery service, and there was a very

657
00:40:34,360 --> 00:40:37,740
suspicious tip very suspicious

658
00:40:37,880 --> 00:40:41,480
amount of tip that was given. Like, you

659
00:40:41,480 --> 00:40:45,160
know, you can imagine, you know, the range of tips can be between x

660
00:40:45,160 --> 00:40:48,605
dollars and y dollars, and suddenly that's, like, you know,

661
00:40:48,605 --> 00:40:51,984
10,000 times y, like, 10,000 times the upper limit.

662
00:40:52,285 --> 00:40:55,964
Yeah. You know, triggers off a suspicious alert. It's

663
00:40:55,964 --> 00:40:59,424
not a normal tip, and it's not a mistake. It's actually, you know, security

664
00:40:59,484 --> 00:41:03,170
issue. So that's an example. Yeah. Interesting. Yeah. I

665
00:41:03,170 --> 00:41:06,930
love the anomaly detection aspect of that. I mean, it just it

666
00:41:07,010 --> 00:41:10,690
it's it's something that we've been doing for a long time,

667
00:41:10,690 --> 00:41:14,495
but then at wrapping it with automation and then

668
00:41:14,495 --> 00:41:17,935
combining that automation with what you just described with all the

669
00:41:17,935 --> 00:41:21,475
agents running down all of the permutations, that

670
00:41:21,615 --> 00:41:25,455
that just sounds amazing. Yeah. It's really cool. I can't

671
00:41:25,455 --> 00:41:28,140
take credit. This isn't me. It's it's it's my team. But,

672
00:41:29,260 --> 00:41:32,859
but I I was like, woah. It's like a hundred bars

673
00:41:32,859 --> 00:41:36,380
running at the same time under one minute. That's amazing. There you go. It's really

674
00:41:36,380 --> 00:41:38,320
cool. Probably smarter than me. But yeah.

675
00:41:40,300 --> 00:41:43,155
That is so awesome. That is cool.

676
00:41:43,875 --> 00:41:47,715
So we we generally have is, we have kind of our

677
00:41:47,715 --> 00:41:51,395
our stock questions that we ask, if you're interested in doing them.

678
00:41:51,395 --> 00:41:54,995
They're not we're not Mike Wallace. We're not trying to I don't even think

679
00:41:54,995 --> 00:41:58,660
anyone gets that reference anymore, but we're not trying to catch you in a,

680
00:41:58,820 --> 00:42:02,420
I gotta come up with a new one, in a thing. But it's mostly, like,

681
00:42:02,420 --> 00:42:05,540
how'd you find your way in the first one is I'll get the rest of

682
00:42:05,540 --> 00:42:09,345
them, up for you in a second. But the first one is, how'd

683
00:42:09,345 --> 00:42:12,865
you find your way into data? Did did the data did you find the data

684
00:42:12,865 --> 00:42:16,385
life or did data life find you? Oh, that's such a great

685
00:42:16,385 --> 00:42:20,165
question. You know, it's funny.

686
00:42:20,625 --> 00:42:24,170
I grew up you know, my my, my mom is a meditation and dance

687
00:42:24,170 --> 00:42:27,450
teacher and my dad is a physics professor. And so,

688
00:42:29,450 --> 00:42:32,329
yeah, and so I, I, you know, grew up with very sort of like, yin

689
00:42:32,329 --> 00:42:34,510
yin yang in my family, if you will.

690
00:42:36,245 --> 00:42:39,605
At a very early age, I used to, like, hang out in in my dad's

691
00:42:39,605 --> 00:42:43,445
lab and, like, do scientific research and stuff like that. So or, you know,

692
00:42:43,445 --> 00:42:46,805
like, very at a very young age, my memories are, like, sitting in a

693
00:42:46,805 --> 00:42:50,470
cinema, watching a movie with my dad and trying to, like, guesstimate how

694
00:42:50,470 --> 00:42:53,990
many people are sitting in the in the audience.

695
00:42:53,990 --> 00:42:57,670
Right? Yes. Just like, you know, I think for, like, a five year

696
00:42:57,670 --> 00:43:01,430
old, it's sort of like a fun fun thing. But, you know, throughout my my

697
00:43:01,430 --> 00:43:05,085
adulthood, like, always sort of had that in in the background. And,

698
00:43:05,484 --> 00:43:09,085
you know, I I think later on in life, I sort of always gravitated towards

699
00:43:09,085 --> 00:43:12,565
data. And when I decided to start a company,

700
00:43:12,565 --> 00:43:16,125
I was actually debating between various areas

701
00:43:16,125 --> 00:43:19,920
like IT and actually blockchain, or, you know,

702
00:43:19,920 --> 00:43:22,800
crypto for a little bit and and data. I think at the end of the

703
00:43:22,800 --> 00:43:26,560
day, like, my heart was really in in data. If I look at, like,

704
00:43:26,560 --> 00:43:30,320
the next ten, twenty years, it's pretty clear to me that data is

705
00:43:30,320 --> 00:43:33,935
gonna be I think it still is the coolest party, and I think it

706
00:43:33,935 --> 00:43:37,555
will be the coolest party to be in. And I personally,

707
00:43:37,695 --> 00:43:41,295
like, you know, it's it's it's funny. Like, throughout my my

708
00:43:41,295 --> 00:43:45,130
career, I've I've also learned the limitations of data. Right? So so data can

709
00:43:45,130 --> 00:43:48,890
tell you whatever story you want. It could tell you, you know, for every question,

710
00:43:48,890 --> 00:43:52,270
it give can give you a yes, and you can also tell a no story.

711
00:43:53,130 --> 00:43:56,970
Right? So so there's also limitations to data,

712
00:43:56,970 --> 00:43:59,655
but but I always have been fascinated,

713
00:44:00,515 --> 00:44:04,115
by by data and space. So can I say both? That's

714
00:44:04,194 --> 00:44:08,035
Yeah. I mean, that's fair. That's fair. Good answer. That's fair. Yep. So

715
00:44:08,035 --> 00:44:10,454
what what's your favorite part of your current job?

716
00:44:12,900 --> 00:44:15,640
Oh, that's hard to choose. I love my job.

717
00:44:16,740 --> 00:44:19,320
I just love it. I think, you know,

718
00:44:20,740 --> 00:44:24,180
the ability to work with customers and actually, like, change the way they

719
00:44:24,180 --> 00:44:27,575
work, I I think that's probably the biggest gratification that I

720
00:44:27,575 --> 00:44:31,255
get, you know, from from my my career. Like, the fact that you can

721
00:44:31,255 --> 00:44:34,935
actually work on something that matters is pretty insane. You know? And when I think

722
00:44:34,935 --> 00:44:38,775
about, like, the future, I'm like, what? So data is gonna be wrong? Like, we're

723
00:44:38,775 --> 00:44:42,400
just gonna be, you know, making decisions off of wrong like, what? I don't

724
00:44:42,400 --> 00:44:46,080
wanna live in that world. You know? And so Yeah. I think

725
00:44:46,080 --> 00:44:49,920
there's something that's, like, really fulfilling and helping, you know, drive a mission that

726
00:44:49,920 --> 00:44:53,360
I believe in that has an impact on customers. And, you know, when customers will

727
00:44:53,360 --> 00:44:57,155
tell me, you know, I started sleeping at night because I

728
00:44:57,155 --> 00:45:00,435
know that, like, I have some coverage for my data. I'm like, yeah. Oh, wow.

729
00:45:00,435 --> 00:45:04,195
I'm glad you're sleeping. You know? Like, good for you. I love

730
00:45:04,195 --> 00:45:08,035
sleeping. So What a cool thing to hear. Yeah. Exactly. I

731
00:45:08,035 --> 00:45:11,680
think that's that's probably, you know, maybe one part. And then the second is, like,

732
00:45:11,680 --> 00:45:15,360
just working with an amazing team. You know, I I spend most of my my

733
00:45:15,360 --> 00:45:18,740
day maybe kinda like, you know, you guys, like, hang out having fun,

734
00:45:18,960 --> 00:45:22,720
laughing. So, you know, I I I'm very

735
00:45:22,720 --> 00:45:26,415
grateful that I get to work with the smartest people on on

736
00:45:26,415 --> 00:45:29,795
worthwhile challenges. Oh, very cool.

737
00:45:30,255 --> 00:45:34,095
We have, three complete these sentences. When I'm not

738
00:45:34,095 --> 00:45:37,395
working, I enjoy blank. Sleeping.

739
00:45:39,640 --> 00:45:43,100
I yeah. I I have a I we recently

740
00:45:43,320 --> 00:45:46,860
have added we we had two kids, and we adopted a cousin. And

741
00:45:47,240 --> 00:45:50,380
I forgot how draining a toddler can be. And I'm

742
00:45:51,595 --> 00:45:54,635
I'm eight to 10 years older since the last time I had a toddler, so

743
00:45:54,635 --> 00:45:58,155
it's like I, I have two

744
00:45:58,155 --> 00:46:01,515
kids, on two under four. So I,

745
00:46:02,635 --> 00:46:06,320
respect the sleep even more. I I can't even I can't

746
00:46:06,320 --> 00:46:10,080
even wrap my head around that. It gets it gets better. I can say

747
00:46:10,080 --> 00:46:13,140
that. It's my own role. I appreciate that.

748
00:46:14,240 --> 00:46:17,460
So our second one is I think the coolest thing in technology

749
00:46:17,680 --> 00:46:21,395
today is blank. The coolest thing in

750
00:46:21,395 --> 00:46:24,995
techno I think the pace of innovation. I think that's really

751
00:46:24,995 --> 00:46:28,435
freaking cool. You know, you can, like, work at a problem today and you're like,

752
00:46:28,515 --> 00:46:32,035
you can't solve this. Two days two days later, a new model will come out.

753
00:46:32,035 --> 00:46:35,390
Boom. You're done. So it's harder. Right? The bar is

754
00:46:35,390 --> 00:46:38,990
higher in order to, like, actually like, it's it's harder to it's

755
00:46:38,990 --> 00:46:42,589
harder to know what to bet on. It's harder to know what the future will

756
00:46:42,589 --> 00:46:45,970
look like, but it's a lot more exciting. So I'm in it.

757
00:46:47,275 --> 00:46:51,035
Cool. Our third and final complete sentence is, I look forward

758
00:46:51,035 --> 00:46:53,775
to the day when I can use technology to blank.

759
00:46:56,075 --> 00:46:59,915
I was always a big fan of teleportation. I think teleportation is really

760
00:46:59,915 --> 00:47:03,420
freaking cool. That would be nice. Can't wait for that. That would be cool.

761
00:47:03,880 --> 00:47:07,000
That would be cool. You know, you're not the first person to answer with them.

762
00:47:07,000 --> 00:47:10,280
Oh, really? Yeah. It's pretty cool. Pretty cool. Sorry.

763
00:47:10,600 --> 00:47:14,200
Number six is share something different about

764
00:47:14,200 --> 00:47:17,224
yourself. Something different.

765
00:47:17,684 --> 00:47:21,525
Yeah. Something different. Let's

766
00:47:21,525 --> 00:47:24,964
see. I mentioned I have two kids. I

767
00:47:24,964 --> 00:47:28,424
meditate when I don't sleep. I like to meditate.

768
00:47:30,390 --> 00:47:34,230
I, what else? I'm married to

769
00:47:34,230 --> 00:47:37,849
my cofounder. Oh, wow. So we,

770
00:47:39,109 --> 00:47:42,789
yeah, we're fortunate to share our lives both at work and at

771
00:47:42,789 --> 00:47:46,525
home. That is cool. Yeah. I can

772
00:47:46,525 --> 00:47:50,285
imagine that would work out really well or not. Like, there's not a lot of

773
00:47:50,285 --> 00:47:54,125
middle ground there. High risk, high reward. High risk, high reward. I

774
00:47:54,125 --> 00:47:57,885
get, like you know, my wife is, you know, she's a

775
00:47:57,885 --> 00:48:01,700
federal employee, and she's, you know, reevaluating what her career

776
00:48:01,700 --> 00:48:05,380
futures look like, you know, and she's like, you

777
00:48:05,380 --> 00:48:09,220
know, I was like, well, you know, you could help. You can start

778
00:48:09,220 --> 00:48:11,540
a new podcast. I can help you with that. She's like, yeah. But then I

779
00:48:11,540 --> 00:48:15,194
have to work with you. And, like, I know what she meant. I know how

780
00:48:15,194 --> 00:48:18,635
it sounds. I know how it sounds, but I know what she means. Like, so

781
00:48:18,635 --> 00:48:22,395
when she did work from home, like, there was literally a, like, an entire floor

782
00:48:22,395 --> 00:48:26,075
between us because Yep. Like, it's too loud. I'm too loud. Yeah. Yeah. Yeah.

783
00:48:26,075 --> 00:48:28,859
Yep. We're very loud too. So

784
00:48:29,720 --> 00:48:33,400
where can folks find more, learn more about, Monte

785
00:48:33,400 --> 00:48:36,460
Carlo and, and and what you're up to?

786
00:48:37,400 --> 00:48:40,875
Probably, I'm the place where I hang out is LinkedIn. So,

787
00:48:41,275 --> 00:48:45,035
I know we just got connected on LinkedIn. That's great. Probably follow me

788
00:48:45,035 --> 00:48:48,155
on LinkedIn or, honestly, reach out to me directly, me,

789
00:48:48,155 --> 00:48:51,915
Moses@MonteCarlodata.com. I hope I don't get a lot of phishing now because

790
00:48:51,915 --> 00:48:55,020
of that. But Well, hopefully, make sure it's the right account because we found out

791
00:48:55,020 --> 00:48:58,480
in the process that there's there was another suspect in

792
00:48:58,540 --> 00:49:02,060
suspicious looking account. And I also think that for our

793
00:49:02,060 --> 00:49:05,740
listeners, it's worth pointing out that I think that people have realized that LinkedIn is

794
00:49:05,740 --> 00:49:09,420
a is a major security vector because I've been getting a lot of

795
00:49:09,420 --> 00:49:13,125
weird a lot more lately. Now I don't think it's related to

796
00:49:13,125 --> 00:49:16,805
the, the refrigerator scandal. Andy and I will do a whole show on that

797
00:49:16,805 --> 00:49:20,505
later because there's there's actually an interesting AI component to that. Okay.

798
00:49:20,565 --> 00:49:24,210
Good to know. And finally, last but not least, Audible

799
00:49:24,210 --> 00:49:27,970
is a sponsor of the podcast. Do you do audiobooks? If

800
00:49:27,970 --> 00:49:30,790
so, recommend one. Otherwise, just recommend a good book you recommend.

801
00:49:32,770 --> 00:49:34,470
A good book. Let's see.

802
00:49:39,025 --> 00:49:41,845
Thinking in bets by Annie Duke.

803
00:49:45,505 --> 00:49:49,285
Professional poker player. Interesting. In in how

804
00:49:51,000 --> 00:49:54,460
lessons from poker can be applied in, in life

805
00:49:55,000 --> 00:49:58,620
and in business. Interesting. I

806
00:49:58,840 --> 00:50:02,680
once worked at a financial services company, and one of the

807
00:50:02,680 --> 00:50:05,995
big shots used to play online poker. And

808
00:50:06,295 --> 00:50:10,055
they're on company, not on company money, but on company time. And a

809
00:50:10,055 --> 00:50:13,515
lot of people Not a lot of people took a dim view of that.

810
00:50:14,535 --> 00:50:18,150
Rightfully so. But he was

811
00:50:18,150 --> 00:50:21,270
making so much money. You know, people that matter didn't take a damn view to

812
00:50:21,270 --> 00:50:24,869
it. When he stopped making so much money, people everyone took a damn view to

813
00:50:24,869 --> 00:50:28,470
it. And it they don't that does end the the story. It

814
00:50:28,470 --> 00:50:32,225
is on I don't see if it's an audio oh, it is

815
00:50:32,225 --> 00:50:35,745
an audio book. It is an audio book. Awesome. I'm gonna add that to my

816
00:50:35,745 --> 00:50:39,505
list. I'm done. Okay. And if you you know, they are a sponsor.

817
00:50:39,505 --> 00:50:43,265
So if you go to, the datadrivenbook.com, you know,

818
00:50:43,265 --> 00:50:47,040
you'll get a free audio book on us. And, you know, if you sign up,

819
00:50:47,040 --> 00:50:50,880
we'll get enough to, you know, buy a coffee.

820
00:50:50,880 --> 00:50:53,940
Maybe not tip them $8,000, but, you know,

821
00:50:54,720 --> 00:50:58,494
we'll get enough for a Starbucks maybe. Maybe. Yeah.

822
00:50:58,494 --> 00:51:01,695
I just tested the link, Frank. Every now and then, we had trouble early on

823
00:51:01,695 --> 00:51:05,375
with the link coming and going. So I just when you saw me turn away

824
00:51:05,375 --> 00:51:09,135
a minute ago when Frank started to this question, that was me typing

825
00:51:09,135 --> 00:51:10,915
in. It worked. It worked.

826
00:51:13,040 --> 00:51:16,880
It's always DNS. That's the Always. It's interesting

827
00:51:16,880 --> 00:51:20,720
you mentioned that. I read an article. Actually, it was a newsletter recently that talked

828
00:51:20,720 --> 00:51:23,859
about, betting being the first stage

829
00:51:24,320 --> 00:51:28,115
in, kind of the path to minimally viable products. And

830
00:51:28,115 --> 00:51:31,954
I thought, now that's curious, and I don't know again, I haven't

831
00:51:31,954 --> 00:51:35,335
read the book. I will listen to it. But the idea of

832
00:51:35,714 --> 00:51:39,335
engaging your team I I manage a team, as well.

833
00:51:39,635 --> 00:51:43,089
And engaging the team by having them do

834
00:51:43,089 --> 00:51:46,770
interesting things and making taking these very large bets

835
00:51:46,770 --> 00:51:48,790
that look nearly impossible,

836
00:51:50,130 --> 00:51:53,730
perhaps. And it's like you said, the the the problem

837
00:51:53,730 --> 00:51:57,505
comes up, and you're thinking this is this is unsolvable. And two days

838
00:51:57,505 --> 00:52:01,105
later, it's solved. And over and over again, I've had that

839
00:52:01,105 --> 00:52:04,785
experience, but I never tied it to the concept of

840
00:52:04,785 --> 00:52:08,385
bets. And I saw this this newsletter that talked about do

841
00:52:08,385 --> 00:52:12,160
that first, And it reminded me a little bit

842
00:52:12,160 --> 00:52:15,680
of Collins talking about, the the big hairy

843
00:52:15,680 --> 00:52:19,360
goals, you know, back in the day. It's very

844
00:52:19,360 --> 00:52:22,480
similar to that maybe in concept. I don't know. I'll have to listen to the

845
00:52:22,480 --> 00:52:26,135
book and check it out, but I was intrigued by the newsletter. Yeah.

846
00:52:26,135 --> 00:52:29,974
There's interesting concepts. Like, I think some of the ideas is, like I mean, even

847
00:52:29,974 --> 00:52:32,695
when you start a company or sort of, you know, start working on a team,

848
00:52:32,695 --> 00:52:36,375
like, you basically have you have a set of cards, which are, like, your strengths,

849
00:52:36,375 --> 00:52:39,870
your weaknesses. And so how how do you play your cards? Like, you can't you

850
00:52:39,870 --> 00:52:43,710
know, if you wanna win around, you can't play with someone else's cards.

851
00:52:43,710 --> 00:52:46,430
You are what you are. And so the best thing you can do is play

852
00:52:46,430 --> 00:52:50,270
with your card. I think that's true for a team solving a problem or startup

853
00:52:50,270 --> 00:52:53,090
or whatever it is. I love that. Yeah.

854
00:52:54,145 --> 00:52:57,745
Interesting. Any final thoughts? This was so fun. Thanks for

855
00:52:57,745 --> 00:53:01,585
having me. Thank you. Thanks for, and you did mention kinda offhand early

856
00:53:01,585 --> 00:53:03,825
on. I don't remember if it was in the green room or not. You have

857
00:53:03,825 --> 00:53:07,460
a podcast yourself? I do not have a podcast myself.

858
00:53:07,520 --> 00:53:11,360
Alright. That was my mistake. Maybe I'll end it tomorrow. Okay. All

859
00:53:11,360 --> 00:53:15,120
good. Life goal one day. There

860
00:53:15,120 --> 00:53:18,960
you go. There you go. And with that, we'll let our AI finish

861
00:53:18,960 --> 00:53:22,595
the show. And that wraps up another data packed episode of

862
00:53:22,595 --> 00:53:26,275
data driven. A massive thank you to our brilliant guest, Bar

863
00:53:26,275 --> 00:53:29,575
Moses, for taking us deep into the world of data observability,

864
00:53:30,195 --> 00:53:33,555
sketchy LinkedIn impersonators, and the dark arts of tipping

865
00:53:33,555 --> 00:53:37,349
anomalies. Who knew a dodgy schema change could cost more than

866
00:53:37,349 --> 00:53:41,109
a luxury sports car? Now, dear listener, if you've made

867
00:53:41,109 --> 00:53:44,950
it this far, you clearly have excellent taste. So why not

868
00:53:44,950 --> 00:53:48,390
put that good judgment to work and leave us a rating and review on

869
00:53:48,390 --> 00:53:51,905
whatever platform you're tuning in on? Apple, Spotify,

870
00:53:52,285 --> 00:53:56,125
Pocket Casts, Morse code, however you get your fix, would love

871
00:53:56,125 --> 00:53:59,825
your feedback. And dare I ask, are you subscribed?

872
00:54:00,525 --> 00:54:04,285
I mean, you wouldn't want to miss out on future episodes filled with more

873
00:54:04,285 --> 00:54:07,530
wit, wisdom, and the occasional fridge based conspiracy,

874
00:54:07,830 --> 00:54:11,349
would you? Until next time, stay curious, stay

875
00:54:11,349 --> 00:54:14,650
observant, and for heaven's sake, keep your data tidy.