Oct. 2, 2024

From Ledgers to Algorithms: Revolutionizing Financial Data

Apple Podcasts podcast player iconSpotify podcast player icon
Apple Podcasts podcast player iconSpotify podcast player icon

Ultimately, providing the best financial advice or service is all about having the right data in the right place at the right time.

This week, host Matt Snow speaks with Chris Gifford, Chief Data and Analytics Officer at USAA Federal Savings Bank, about the evolution of data analytics within the financial industry—as well as the importance of having a strong data foundation. Chris also shares insights on driving innovation through real-world problem-solving, the democratization of data, and the need to break down data silos for seamless integration.

Join us as we discuss:

  • The democratization of deep data analysis and its effects
  • What factors organizations should consider when deciding whether to build, buy, or partner for technological solutions
  • Using timely data analysis to give personalized advice to members
  • Strategies for fostering a culture of innovation within data analytics teams
WEBVTT

1
00:00:02.879 --> 00:00:05.639
You are listening to Leaders in Lending from Upstart, a

2
00:00:05.679 --> 00:00:09.560
podcast dedicated to helping consumer lenders grow their programs and

3
00:00:09.640 --> 00:00:13.919
improve their product offerings. Each week, here, decision makers in

4
00:00:13.960 --> 00:00:16.760
the finance industry offer insights into the future of the

5
00:00:16.800 --> 00:00:22.039
lending industry, best practices around digital transformation, and more. Let's

6
00:00:22.039 --> 00:00:22.800
get into the show.

7
00:00:26.760 --> 00:00:28.719
Welcome to another episode of Leaders in Lending.

8
00:00:28.760 --> 00:00:31.679
I'm your host Matt Snow today joined by Chris Gifford,

9
00:00:31.800 --> 00:00:34.200
chief Data and Analytics Officer from USAA.

10
00:00:34.320 --> 00:00:36.000
Chris, Welcome, Thank you, Matt.

11
00:00:36.039 --> 00:00:37.560
Good to see you again. It's been a while.

12
00:00:38.119 --> 00:00:40.119
Yeah, good to see you. Always good to catch up.

13
00:00:40.159 --> 00:00:42.079
I know, you know, it's been a while since we've

14
00:00:42.079 --> 00:00:44.799
worked together, way back in the day. A lot's happened

15
00:00:44.840 --> 00:00:47.520
in data and analytics since then, so I'm really glad

16
00:00:47.560 --> 00:00:49.359
to be able to have some time to dig into

17
00:00:49.359 --> 00:00:49.679
that with you.

18
00:00:50.399 --> 00:00:53.560
Yeah, I'm back, we're working together, Chase. I don't think

19
00:00:53.560 --> 00:00:55.119
our hair was gray, and we might have had more

20
00:00:55.119 --> 00:00:57.759
of it. So it's been a while, but always good

21
00:00:57.799 --> 00:01:00.679
to check in. And what an exciting and field that

22
00:01:00.679 --> 00:01:04.079
we're talking about too, and so much has changed it's

23
00:01:04.120 --> 00:01:05.120
hard to even keep up.

24
00:01:05.000 --> 00:01:07.560
With It's true and you know, I was doing some

25
00:01:07.760 --> 00:01:10.000
research and preparation for our conversation.

26
00:01:10.280 --> 00:01:11.519
I wasn't sure i'd find anything.

27
00:01:11.519 --> 00:01:14.959
But do you know who and when the first chief

28
00:01:15.040 --> 00:01:19.200
data officer could have came about? No, well, you know,

29
00:01:19.319 --> 00:01:22.640
unsurprisingly it's in banking, so it was actually Capital One.

30
00:01:22.799 --> 00:01:26.040
Catherine clay daw seems to be referenced as the first

31
00:01:26.200 --> 00:01:29.000
Cheetah Chief Data Officer back in two thousand and two,

32
00:01:29.079 --> 00:01:31.799
so you know around that time we were together.

33
00:01:31.959 --> 00:01:34.640
Wow, that is interesting. I hadn't I hadn't looked that up,

34
00:01:34.719 --> 00:01:37.959
so I knew it's been around relatively short period of time,

35
00:01:38.000 --> 00:01:42.280
but very interesting. Cap One makes sense. They do a

36
00:01:42.280 --> 00:01:43.319
lot with data over there.

37
00:01:43.640 --> 00:01:44.200
Absolutely.

38
00:01:44.239 --> 00:01:46.840
Well, maybe we can start with that, maybe give a

39
00:01:46.879 --> 00:01:48.719
little bit of an overview of what, you know, the

40
00:01:48.799 --> 00:01:51.239
chief data analytics officer role looks like at USA.

41
00:01:51.319 --> 00:01:52.879
How did you come to that, and a little bit

42
00:01:52.879 --> 00:01:53.480
of your journey.

43
00:01:54.079 --> 00:01:57.200
Yeah, so I'll take you back, not all the way

44
00:01:57.239 --> 00:02:00.000
to the very beginning. But as I was finishing school

45
00:02:00.239 --> 00:02:02.000
and looking for my first job, I took a job

46
00:02:02.040 --> 00:02:05.599
at Gie Capital answering the phones. You know, I think

47
00:02:05.640 --> 00:02:08.199
you for calling Monogram Bank USA. This is Chris Gifford

48
00:02:08.199 --> 00:02:10.960
and so it's been a journey since then, but that

49
00:02:11.080 --> 00:02:13.400
was my very first job out of college. And while

50
00:02:13.400 --> 00:02:15.840
I was sitting there, I noticed that this terminal that

51
00:02:15.960 --> 00:02:17.840
was on the desk in front of me seemed to

52
00:02:17.840 --> 00:02:20.479
have capabilities. So I asked for the manual to it

53
00:02:20.960 --> 00:02:25.080
and started writing macros to do my job because I

54
00:02:25.199 --> 00:02:27.560
was lazy, but had the ability to try to figure

55
00:02:27.560 --> 00:02:29.000
out how to make the machine do it for me,

56
00:02:29.520 --> 00:02:31.879
and kind of launched into this journey of what we

57
00:02:31.879 --> 00:02:35.280
would today call robotic process automation. And I wrote a

58
00:02:35.319 --> 00:02:38.439
significant string of macros that did a bunch of my job,

59
00:02:38.520 --> 00:02:41.000
and my boss kind of tapped me on the shoulder

60
00:02:41.000 --> 00:02:45.039
and said, hey, why are your stats thirty five percent

61
00:02:45.080 --> 00:02:46.919
better than everyone else on the floor and you have

62
00:02:47.039 --> 00:02:49.560
like zero errors? And so I showed them what I

63
00:02:49.639 --> 00:02:52.520
was doing and they said, okay, fantastic, you do that.

64
00:02:52.639 --> 00:02:54.560
Now you're off the phones. And that was kind of

65
00:02:54.560 --> 00:02:58.800
the first launch into data and analytics, computers, computer aid

66
00:03:00.159 --> 00:03:04.800
or assisted even digital processing, robotic process automation, and just

67
00:03:04.879 --> 00:03:09.120
automation in general. So that started Capital. They have a

68
00:03:09.199 --> 00:03:12.039
great program that they identify people with you know, skills

69
00:03:12.159 --> 00:03:14.639
or the desire they will train them and so They

70
00:03:14.639 --> 00:03:17.360
put me through a lot of training, and I went

71
00:03:17.400 --> 00:03:21.080
from there into adaptive control systems, building models back on

72
00:03:21.120 --> 00:03:24.120
the mainframe. I didn't use punch cards. I'm not that old,

73
00:03:24.840 --> 00:03:27.159
but I know a lot of people that did. But no,

74
00:03:27.319 --> 00:03:30.599
we were writing code on the mainframe and a lot

75
00:03:30.639 --> 00:03:33.639
in SaaS, but building algorithms and using those in our

76
00:03:33.680 --> 00:03:38.199
adaptive control system strategies to drive you know, credit card

77
00:03:38.360 --> 00:03:44.120
line management, originations, authorizations, reissue, those sorts of things. From there,

78
00:03:44.919 --> 00:03:48.680
I went to Nation's Bank down in Charlotte, North Carolina,

79
00:03:48.719 --> 00:03:51.199
Carolina for a bit. This was prior to the b

80
00:03:51.280 --> 00:03:53.000
of a merger, doing more of the same but a

81
00:03:53.000 --> 00:03:57.479
bigger portfolio, and then moved to Bank one in Columbus, Ohio,

82
00:03:57.520 --> 00:03:59.960
and survived all of those mergers up through and included

83
00:04:00.759 --> 00:04:02.879
Chase at the end of it. I spent eighteen years

84
00:04:02.879 --> 00:04:04.639
at Chase doing more of the same. As I was

85
00:04:04.800 --> 00:04:08.479
leaving Chase, I was head of customer Data and Analytics

86
00:04:08.520 --> 00:04:12.560
and really building out that full customer three sixty bringing

87
00:04:12.599 --> 00:04:16.079
online data products for consumption in various use cases, and

88
00:04:16.160 --> 00:04:19.879
really trying to advance the use of modeling into all

89
00:04:19.920 --> 00:04:23.639
the decisions and optimization engines. So from a career standpoint.

90
00:04:23.959 --> 00:04:26.920
That's right at the last minute before I made the

91
00:04:27.000 --> 00:04:28.399
jump over to USAA.

92
00:04:29.680 --> 00:04:31.759
Gotcha, And so what are some of your top priorities

93
00:04:31.800 --> 00:04:34.199
at USAA, Like, what are the challenges there that you

94
00:04:34.199 --> 00:04:36.680
guys are trying to solve with data and analytics?

95
00:04:37.040 --> 00:04:41.199
Yeah? Sure, So when I joined USAA, they weren't as

96
00:04:41.279 --> 00:04:43.519
far along, I guess in their data journey as what

97
00:04:43.600 --> 00:04:45.600
Chase was. As you could imagine Chase being a very,

98
00:04:45.680 --> 00:04:49.120
very large financial institution. So step one really was build

99
00:04:49.160 --> 00:04:52.439
a foundation of data, and it is get all of

100
00:04:52.480 --> 00:04:58.600
the data, get it cleaned, hygiened available. We kind of

101
00:04:58.600 --> 00:05:02.360
skipped through some of the old are evolutions of data

102
00:05:02.439 --> 00:05:04.560
and data warehousing. We didn't go to kind of the

103
00:05:04.560 --> 00:05:07.839
big industry third normal form warehouses, went more to a

104
00:05:07.920 --> 00:05:12.160
modern lakehouse architecture, and here most recently rehosted the entire

105
00:05:12.199 --> 00:05:16.240
analytics environment up on the cloud, which where we can

106
00:05:16.639 --> 00:05:21.600
use entire lines of code to expand the universe, increase

107
00:05:21.680 --> 00:05:26.639
compute nodes, bring in other data sources. We're now working

108
00:05:26.680 --> 00:05:30.199
on implementing cloud native tools so we can do jobs

109
00:05:30.279 --> 00:05:32.600
end to end and stay in the cloud the entire time.

110
00:05:32.680 --> 00:05:37.000
Really trying to press for speed and ease in the use.

111
00:05:37.560 --> 00:05:39.680
I think the other part that we've been working on

112
00:05:39.920 --> 00:05:42.839
is managing risk. Right, with a lot of data comes

113
00:05:42.839 --> 00:05:44.600
a lot of risk, so you have to protect it.

114
00:05:44.959 --> 00:05:47.720
There is sensitive data p I I, pc I data

115
00:05:48.519 --> 00:05:52.279
and lots of evolving laws, rules and rights around how

116
00:05:52.360 --> 00:05:54.759
you manage it and protected data for your we call

117
00:05:54.759 --> 00:05:58.040
the members, but for your customers. I think the other

118
00:05:58.120 --> 00:06:03.120
thing is you need folks on your team that have

119
00:06:03.279 --> 00:06:05.720
the ability to understand how does the business work, how

120
00:06:05.759 --> 00:06:08.480
does the business make money, how does a bank make money?

121
00:06:08.839 --> 00:06:11.279
So it would be nice, you know, some days I

122
00:06:11.360 --> 00:06:13.360
wish I was just still sitting down in the basement

123
00:06:13.480 --> 00:06:16.000
of a bank somewhere in a corner, writing code and

124
00:06:16.040 --> 00:06:19.160
not talking to anyone. But really, for us to find

125
00:06:19.160 --> 00:06:22.079
the value in data, we have to understand the business,

126
00:06:22.279 --> 00:06:25.920
build those business relationships with our product owners, our process owners,

127
00:06:25.920 --> 00:06:29.279
our channel owners, and figure out if we bring data

128
00:06:29.360 --> 00:06:32.800
to the problems that they're experiencing, what kind of value

129
00:06:33.360 --> 00:06:38.360
and opportunities exist there. And so I think it's that piece.

130
00:06:39.959 --> 00:06:42.839
Building those bridges and being able to talk the language

131
00:06:42.839 --> 00:06:46.920
of business, educating your business partners on data in its

132
00:06:46.959 --> 00:06:49.279
value and how can you use it? Is kind of

133
00:06:49.279 --> 00:06:51.480
a key part of the chief data and analytics officers,

134
00:06:51.480 --> 00:06:54.279
probably not as well talked about as you know, the

135
00:06:54.360 --> 00:06:57.079
modeling and how many petabytes of data do you have

136
00:06:57.240 --> 00:06:58.920
and what platform are you on and what tools do

137
00:06:58.959 --> 00:07:03.040
you use, but building those just foundational relationships with the

138
00:07:03.079 --> 00:07:06.600
business to understand what their main problems are is really critical.

139
00:07:07.079 --> 00:07:09.439
One last piece kind of the secret sauce maybe to

140
00:07:09.480 --> 00:07:12.439
all of this is you have to constantly drive innovation

141
00:07:13.439 --> 00:07:16.560
this area, just like we talked about earlier, this area

142
00:07:16.639 --> 00:07:20.879
of data and algorithms and data science and even generative AI.

143
00:07:20.920 --> 00:07:23.759
I can't go through a podcast without saying generative AI, right,

144
00:07:24.680 --> 00:07:27.279
But even these areas that are evolving so fast, you

145
00:07:27.360 --> 00:07:30.959
have to constantly be pushing the envelope on innovation. And

146
00:07:31.000 --> 00:07:33.959
so some of the things that we drive here your

147
00:07:34.040 --> 00:07:37.160
standard hackathons. We've done hackathons what we call it like

148
00:07:37.160 --> 00:07:39.959
speed dating. We bring a business people in, we set

149
00:07:40.000 --> 00:07:42.480
them at a table, and we send a group of

150
00:07:42.560 --> 00:07:44.800
data scientists at them, and they've got two minutes to

151
00:07:44.839 --> 00:07:48.040
describe their business problem. And then the data scientists team

152
00:07:48.079 --> 00:07:50.720
go table to table to all the business people and

153
00:07:50.839 --> 00:07:54.199
they choose, Okay, which problem do I want to try

154
00:07:54.240 --> 00:07:57.040
to solve over the next two days, and we invest

155
00:07:57.120 --> 00:07:59.199
a lot into these. The teams absolutely love them, and

156
00:07:59.240 --> 00:08:01.639
I don't think we've had one yet where we didn't

157
00:08:01.680 --> 00:08:05.160
at least get multiple patents and you know, multiple millions

158
00:08:05.199 --> 00:08:08.399
of dollars of business value. So it just continues to

159
00:08:08.519 --> 00:08:12.360
press uh, press us forward to keep trying to stay

160
00:08:12.360 --> 00:08:15.000
on the cutting edge and UH and use all these

161
00:08:15.040 --> 00:08:17.879
tools and all this data for for real business value.

162
00:08:18.160 --> 00:08:20.199
Yeah, awesome. I love a lot in there.

163
00:08:20.240 --> 00:08:23.360
And you did beat me to the AI Bingo card first,

164
00:08:23.839 --> 00:08:24.959
So maybe we can come back.

165
00:08:24.879 --> 00:08:27.319
To innovation as a topic. But you know, you hit

166
00:08:27.360 --> 00:08:27.720
on something.

167
00:08:27.720 --> 00:08:29.480
I want to get more of your perspective on the

168
00:08:29.519 --> 00:08:33.039
idea of getting value out of the data or aligning

169
00:08:33.120 --> 00:08:35.559
back to the business, because I think we are in

170
00:08:35.600 --> 00:08:37.600
a bit of a maybe a swing of the pendulum

171
00:08:37.679 --> 00:08:41.759
here where the focus is less on explainability or understanding

172
00:08:42.120 --> 00:08:45.159
how some of these technology work, like ll M or

173
00:08:45.159 --> 00:08:47.600
some of the generative AI. Like it just works and

174
00:08:47.639 --> 00:08:51.000
it's cool and it's fun right now, but do you

175
00:08:51.080 --> 00:08:53.519
see like different use cases where things should be a

176
00:08:53.519 --> 00:08:56.840
lot more rigorous around the explanation and like the tie

177
00:08:56.919 --> 00:09:00.679
to value from data to the analytics the insight back

178
00:09:00.679 --> 00:09:04.159
to the business, or is there is there a use

179
00:09:04.200 --> 00:09:06.000
in a world where it's like, hey, it just works,

180
00:09:06.000 --> 00:09:07.559
so let's roll with it and onto the next thing,

181
00:09:07.600 --> 00:09:08.720
like how do you think about.

182
00:09:09.399 --> 00:09:13.720
Yeah, there's there's definitely both an evolution and two categories

183
00:09:13.759 --> 00:09:15.960
of kinds of models and problems that you kind of solve.

184
00:09:16.039 --> 00:09:18.840
So so the start with the evolution of it. I

185
00:09:19.000 --> 00:09:22.759
can remember within my career having a I won't say

186
00:09:22.799 --> 00:09:25.240
which company, even though I've already named the wall, but

187
00:09:25.360 --> 00:09:28.759
somewhere back along that line, I had a boss that said,

188
00:09:28.759 --> 00:09:31.600
I don't believe in models. I think, you know, we

189
00:09:31.639 --> 00:09:34.919
should underwrite loans by looking at their income and having

190
00:09:34.919 --> 00:09:40.360
a person judgmentally decide this or that. And so that

191
00:09:40.360 --> 00:09:44.600
that was again not that that's within my career event horizon, right,

192
00:09:44.679 --> 00:09:47.480
so we're not that far removed from it, but we

193
00:09:47.600 --> 00:09:50.879
have come a long way now into saying no, we

194
00:09:50.919 --> 00:09:53.960
can use models to help us make better decisions, even

195
00:09:54.000 --> 00:09:57.679
help us avoid some of the maybe subconscious biases that

196
00:09:57.720 --> 00:09:59.720
we as humans have when we look at data or

197
00:09:59.759 --> 00:10:02.559
make decisions like that. And I think that's where you

198
00:10:02.639 --> 00:10:04.279
have to pair it up with the laws, rules and

199
00:10:04.320 --> 00:10:06.480
REGs in this space, which really are put in place

200
00:10:06.519 --> 00:10:10.799
to help us not put in or allow those biases,

201
00:10:11.039 --> 00:10:15.320
or you know, have people being impacted disparately across these

202
00:10:15.320 --> 00:10:17.759
different decision areas. So to kind of put a finer

203
00:10:17.799 --> 00:10:20.320
point on that, like a credit risk decision, should I

204
00:10:20.440 --> 00:10:24.879
extend credit to someone? Yes or no? Is one where yes,

205
00:10:24.919 --> 00:10:28.120
there are opportunities for models, but it's probably not going

206
00:10:28.159 --> 00:10:31.360
to be a oh it's a black box. The algorithms

207
00:10:31.399 --> 00:10:34.279
always right, just trust it, right, We have to be

208
00:10:34.360 --> 00:10:37.200
able to show what's going on inside of the model.

209
00:10:37.679 --> 00:10:40.320
Once you get beyond you know, two or three way

210
00:10:40.360 --> 00:10:43.399
interactions of data variables, the brain starts going like WHOA,

211
00:10:43.639 --> 00:10:46.120
I don't understand this anymore. And some of these algorithms

212
00:10:46.200 --> 00:10:49.960
can have sixteen, eighteen twenty way interactions of variables where

213
00:10:50.000 --> 00:10:53.159
we have no Our brain has no ability to comprehend

214
00:10:53.200 --> 00:10:55.960
how is it getting to its answer? But yet we

215
00:10:56.039 --> 00:10:59.320
can show through additional add on analysis, are is it

216
00:10:59.360 --> 00:11:01.840
doing the right thing? Is it making the right decisions?

217
00:11:02.120 --> 00:11:05.600
Are there disparate impacts or their unintended consequences of the model,

218
00:11:05.840 --> 00:11:09.600
and those things need to be layered on and so again,

219
00:11:09.639 --> 00:11:12.240
the credit risk decision is a key one Obviously, fair

220
00:11:12.279 --> 00:11:16.519
Credit Reporting Act and available availability of credit to all

221
00:11:17.039 --> 00:11:21.879
are key reasons why and necessary so that we have

222
00:11:21.960 --> 00:11:27.000
a good, solid and fair financial services industry. Some of

223
00:11:27.000 --> 00:11:30.519
the other areas, though, like who should we send a

224
00:11:30.799 --> 00:11:34.279
marketing mailer to? Actually, you and I used to do

225
00:11:34.320 --> 00:11:35.759
mailers back in the day, right, I don't even think

226
00:11:35.759 --> 00:11:37.960
they send out mail anymore. But maybe it's emails, right,

227
00:11:38.080 --> 00:11:41.000
or alerts or those kinds of things. So if you

228
00:11:41.240 --> 00:11:43.879
send an alert, it's like, hey, are you interested in

229
00:11:44.480 --> 00:11:47.919
applying for a whatever? Or Hey, we've got a partnership

230
00:11:47.919 --> 00:11:49.799
with this other company, would you like to see some

231
00:11:49.879 --> 00:11:54.679
of their things? Certainly it can be annoying to a member,

232
00:11:54.679 --> 00:11:56.840
but it doesn't really put them at risk. It doesn't

233
00:11:56.840 --> 00:11:59.080
put the company at risk. So there are some more

234
00:11:59.120 --> 00:12:02.960
innocuous of things that we've can be doing with with

235
00:12:03.080 --> 00:12:08.080
the models and the AI to target and optimize in

236
00:12:08.159 --> 00:12:10.679
ways that we can't really see or understand. Maybe another

237
00:12:10.679 --> 00:12:14.480
one would be trying to figure out why is this

238
00:12:14.559 --> 00:12:19.000
customer calling and automatically wrap them to a customer service

239
00:12:19.039 --> 00:12:21.480
rep that is best suited to handle that kind of

240
00:12:21.559 --> 00:12:24.759
a problem, so again not really harming the customer, trying

241
00:12:24.759 --> 00:12:28.639
to drive efficiency some of just those back office operational

242
00:12:28.799 --> 00:12:31.600
you know fish where the fish are kinds of things

243
00:12:32.039 --> 00:12:34.360
are certainly there. And then when you start thinking about

244
00:12:34.399 --> 00:12:38.799
like the cool new kinds of generative AI solutions, man,

245
00:12:38.840 --> 00:12:41.120
I get excited. I want to be able to ask

246
00:12:41.240 --> 00:12:45.840
my supercharged digital assistance, like hey, go get me new

247
00:12:45.840 --> 00:12:48.679
homeowners insurance. And I wanted to like to be able

248
00:12:48.720 --> 00:12:51.360
to go do that, right, And so you start getting

249
00:12:51.360 --> 00:12:54.720
into could could I could we as a bank? Could

250
00:12:54.759 --> 00:12:56.720
we as a country? Could the government allow us to

251
00:12:56.759 --> 00:12:59.559
get to a point where we have digital assistance like

252
00:12:59.600 --> 00:13:02.120
that that have the free reign to go out and

253
00:13:02.720 --> 00:13:05.799
solve these kinds of problems for us? Certainly it's easy.

254
00:13:05.799 --> 00:13:09.279
Now It's like, hey, here's the entire year end financial

255
00:13:09.320 --> 00:13:12.799
report for whatever company. Can you distill this down to

256
00:13:12.840 --> 00:13:15.399
the key points and compare to you know, their competitors.

257
00:13:15.519 --> 00:13:19.200
That kind of stuff is available, these ministryvia kinds of

258
00:13:19.240 --> 00:13:21.799
things I want to be handled by AI and m

259
00:13:21.919 --> 00:13:24.480
L And so I think in our personal lives and

260
00:13:24.639 --> 00:13:28.759
areas outside heavily regulated businesses, there's probably more opportunities for

261
00:13:28.799 --> 00:13:32.840
it to serve us and create experiences that are very cool.

262
00:13:33.600 --> 00:13:39.240
Within financial services obviously, and necessarily a highly regulated, regulated business.

263
00:13:39.240 --> 00:13:43.080
There there's a lot more concern and caution and controls

264
00:13:43.080 --> 00:13:45.240
we need to rep around it. Sorry, that was a

265
00:13:45.279 --> 00:13:48.080
long answer, did I did? I covered a lot.

266
00:13:48.279 --> 00:13:48.919
No, I like it.

267
00:13:49.000 --> 00:13:51.120
I think, you know, one of the other things that

268
00:13:51.279 --> 00:13:54.159
there's again lots of like similarities and being aware of

269
00:13:54.200 --> 00:13:57.240
the history of like you know, even with the more

270
00:13:57.279 --> 00:13:59.639
complicated models, maybe it's even more true like the garbage

271
00:13:59.679 --> 00:14:02.159
and guard the job right, So you talked about generative

272
00:14:02.159 --> 00:14:04.519
AI and some of those, like you know, understanding the

273
00:14:04.600 --> 00:14:08.799
data that these algorithms and machines are using to base

274
00:14:08.879 --> 00:14:14.039
their recommendations or insights on is important. Even some of

275
00:14:13.840 --> 00:14:17.279
the changes in innovation, I'm seeing more of being able

276
00:14:17.320 --> 00:14:20.080
to provide references or links like here's why I'm giving

277
00:14:20.159 --> 00:14:22.559
you this answer, here's you know, maybe a study I've

278
00:14:22.600 --> 00:14:27.120
seen or other sites. So I do, like, like you said,

279
00:14:27.440 --> 00:14:29.960
really excited about the space and the rate of change

280
00:14:29.960 --> 00:14:34.720
and the ability to add you know, more insight and

281
00:14:34.799 --> 00:14:37.080
speed to I think call back to what you were saying,

282
00:14:37.120 --> 00:14:42.279
like you know, the the usefulness of these technologies really

283
00:14:42.279 --> 00:14:45.200
comes back to how individuals or members are going to

284
00:14:45.200 --> 00:14:49.600
get value out of them that we provide. So in

285
00:14:49.679 --> 00:14:52.320
terms of other innovations, the hack of phone, I wanted

286
00:14:52.320 --> 00:14:53.759
to come back to that because I really like the

287
00:14:53.799 --> 00:14:55.399
idea of combining the.

288
00:14:55.360 --> 00:14:56.399
Business in that.

289
00:14:56.519 --> 00:14:58.840
You know, a lot of the hackathons I see kind

290
00:14:58.879 --> 00:15:01.480
of come more ground up, you know, engineers or programmers

291
00:15:01.480 --> 00:15:04.200
thinking about ways to make maybe their lives easier, or

292
00:15:04.360 --> 00:15:08.879
some consumer focus, but usually generated from within those units.

293
00:15:09.679 --> 00:15:11.279
Was that something that was always a part of your

294
00:15:11.279 --> 00:15:15.120
guys program to include the business or I like that combination?

295
00:15:15.639 --> 00:15:17.879
Yeah, and it's something that's strilled into me. I had

296
00:15:17.879 --> 00:15:20.320
a really, really good manager at Chase, and again I

297
00:15:20.360 --> 00:15:22.320
won't name names because I've had a lot of good

298
00:15:22.360 --> 00:15:24.159
managers through the years, but I can still hear his

299
00:15:24.440 --> 00:15:28.240
voice saying, solve business problems, right. That's That's what the

300
00:15:28.320 --> 00:15:31.720
data analytics team was there to do. It wasn't to advance,

301
00:15:31.799 --> 00:15:35.200
you know, the capabilities of science. We weren't doing science

302
00:15:35.240 --> 00:15:38.639
fare projects. We were there to solve real business problems.

303
00:15:38.679 --> 00:15:41.039
And that's something I've carried forward in my career and

304
00:15:41.080 --> 00:15:44.159
I've really kind of preached that into my teams, Like

305
00:15:44.200 --> 00:15:47.159
I will do hackathons I'm doing monthly, I'll do them quarterly.

306
00:15:47.200 --> 00:15:50.639
I'll give out large prizes if you win them, but

307
00:15:50.679 --> 00:15:53.039
we're going to work on solving business problems if we're

308
00:15:53.039 --> 00:15:55.240
going to take them on. Not to say that there

309
00:15:55.240 --> 00:15:59.879
aren't opportunities in those back office roles in the data engineering,

310
00:16:00.440 --> 00:16:06.120
especially with some of the generative AI auto code kinds

311
00:16:06.159 --> 00:16:08.279
of things like you can think about in Get Up

312
00:16:08.480 --> 00:16:11.039
or you know, I've watched some of my team go

313
00:16:11.120 --> 00:16:13.000
from you know, writing code long hand, all of a

314
00:16:13.039 --> 00:16:15.480
sudden it's just like they're just hitting their hitting tap right,

315
00:16:15.559 --> 00:16:17.559
and it's just it's like writing a line of code

316
00:16:17.600 --> 00:16:19.720
at a time and they're just kind of overseeing it

317
00:16:19.759 --> 00:16:21.919
to make sure it all works. So certainly there are

318
00:16:22.799 --> 00:16:26.039
a large number of opportunities in the back office spaces

319
00:16:26.039 --> 00:16:30.080
for these kinds of things to have value. But you know, again,

320
00:16:30.120 --> 00:16:31.679
at the end of the day, I think it is

321
00:16:32.600 --> 00:16:35.360
at USAA and kind of our focus on our mission

322
00:16:35.440 --> 00:16:38.879
and our in serving our members, it's really about what's

323
00:16:38.919 --> 00:16:41.639
one more thing that we can do to take problems

324
00:16:41.840 --> 00:16:44.480
off of our members plate and help them, you know,

325
00:16:44.960 --> 00:16:47.559
achieve financial stability in their lives.

326
00:16:47.840 --> 00:16:49.919
Yeah, totally agree. I like that approach.

327
00:16:51.519 --> 00:16:54.080
Hey, there, former host Jeff calviner here to let you

328
00:16:54.120 --> 00:16:57.399
know about an exciting opportunity to strengthen your understanding of

329
00:16:57.519 --> 00:17:01.000
artificial intelligence and financial services. If you want to better

330
00:17:01.039 --> 00:17:03.840
understand the fundamentals of how AI models work and how

331
00:17:03.840 --> 00:17:07.680
they will impact the financial services industry, your luck. Upstart

332
00:17:07.720 --> 00:17:10.440
is launched an online course and certification designed to do

333
00:17:10.720 --> 00:17:13.160
just that. This course will give you the knowledge you

334
00:17:13.240 --> 00:17:16.240
need to leave the conversation about how your organization can

335
00:17:16.319 --> 00:17:20.759
best leverage and manage this revolutionary technology. If that sounds interesting,

336
00:17:21.160 --> 00:17:23.319
check out the course as it's currently free for a

337
00:17:23.359 --> 00:17:26.240
limited time. You can learn more at upstart dot com

338
00:17:26.279 --> 00:17:31.279
slash AI Certification. Again, that's upstart dot Com slash AI Certification.

339
00:17:31.799 --> 00:17:33.680
Thanks, and now back to the show.

340
00:17:35.319 --> 00:17:37.720
What other innovations have you seen in the past, you know,

341
00:17:37.759 --> 00:17:42.359
a couple of decades in terms of analytics, data insights,

342
00:17:43.240 --> 00:17:46.000
just that whole evolution within in that space.

343
00:17:46.960 --> 00:17:49.440
Yeah, the I can remember kind of in my career.

344
00:17:50.680 --> 00:17:54.359
Again I won't say where, but one day we first

345
00:17:54.400 --> 00:17:58.559
built the data warehouse, right and before that, I think

346
00:17:58.599 --> 00:18:02.799
everybody in the world was using what different names across

347
00:18:02.799 --> 00:18:04.759
different companies. But it's like some kind of a marketing

348
00:18:04.799 --> 00:18:08.160
extract file em for you know, So the hey, we

349
00:18:08.559 --> 00:18:10.480
pulled a file and you guys can start using it.

350
00:18:10.559 --> 00:18:13.039
And we were doing analysis and we were creating marketing

351
00:18:13.079 --> 00:18:15.920
files and those kinds of things, and then we started

352
00:18:15.960 --> 00:18:19.319
loading those into a warehouse. And I can remember we

353
00:18:19.400 --> 00:18:21.799
loaded the very first month of data into the warehouse

354
00:18:21.839 --> 00:18:25.079
and was like, all right, fantastic. It took four months

355
00:18:25.079 --> 00:18:27.279
to get the second month of data into the warehouse.

356
00:18:27.319 --> 00:18:29.720
But all of a sudden, we had like time series data,

357
00:18:29.759 --> 00:18:33.279
and there's a whole new era of analyses that we

358
00:18:33.319 --> 00:18:35.720
could start doing. And when you start loading you know,

359
00:18:35.759 --> 00:18:40.759
those and constantly well defined structured data with quality, start

360
00:18:40.759 --> 00:18:43.319
putting that time series together, all of a sudden, you

361
00:18:43.359 --> 00:18:46.599
can just bring data to problems like here's how many

362
00:18:46.640 --> 00:18:51.079
widgets you do per time period, and how many have errors,

363
00:18:51.400 --> 00:18:55.359
and just this entire amount of analysis that starts just

364
00:18:55.440 --> 00:18:59.359
bringing insights into problems. From there you go onto building

365
00:18:59.400 --> 00:19:04.640
models and optimizing and then auto optimizing these different things.

366
00:19:04.720 --> 00:19:07.640
And so that evolution has been very interesting to watch.

367
00:19:07.960 --> 00:19:10.160
I think in the modeling space, you know, started out

368
00:19:10.200 --> 00:19:15.079
just with straightforward old linear regression, went to logistic regression,

369
00:19:15.400 --> 00:19:17.920
and then kind of at a point in time, it

370
00:19:18.039 --> 00:19:21.920
just seemed like this entire world of models and algorithms

371
00:19:21.960 --> 00:19:24.720
exploded on what was available to us. And next thing

372
00:19:24.720 --> 00:19:29.039
you know, we've got random forest and genetic models and

373
00:19:29.160 --> 00:19:32.799
just all kinds of things that increased the predictive power

374
00:19:32.880 --> 00:19:35.400
in these spaces. Now again that's where we started getting

375
00:19:35.440 --> 00:19:37.960
into But can you explain it? She had to be

376
00:19:38.000 --> 00:19:41.200
careful which algorithm you used, But certainly you know, over

377
00:19:41.240 --> 00:19:44.720
the last man twenty five thirty years, we've seen an

378
00:19:44.759 --> 00:19:49.519
explosion of those kinds of models and algorithms available to us.

379
00:19:49.599 --> 00:19:54.480
The other thing I would say, I'm not a statistician.

380
00:19:54.880 --> 00:19:57.319
I don't have a degree in applied math and stats.

381
00:19:58.119 --> 00:20:00.839
But what I did was I learned that the tools

382
00:20:01.519 --> 00:20:04.440
that these algorithms come in make it easy for a

383
00:20:04.480 --> 00:20:07.640
person even like myself, to be able to build models

384
00:20:07.680 --> 00:20:10.799
and deploy them and add value into the business. So

385
00:20:10.880 --> 00:20:14.960
the enablement of tools and just a significant increase of data,

386
00:20:15.079 --> 00:20:20.799
data storage, and compute horsepower has made modeling accessible to

387
00:20:20.960 --> 00:20:24.480
a much broader group of people. Certainly, I have tremendous

388
00:20:24.480 --> 00:20:29.839
respect and for true applied math, stats PhDs and majors,

389
00:20:29.880 --> 00:20:32.279
and I hire a ton of them. But I'm just

390
00:20:32.279 --> 00:20:34.559
saying there more and more people can play in the space.

391
00:20:34.640 --> 00:20:37.759
Given the advent of those tools as well.

392
00:20:38.279 --> 00:20:39.799
Yeah, I couldn't agree more.

393
00:20:39.839 --> 00:20:41.559
And I haven't gotten to the point where like I'm

394
00:20:41.599 --> 00:20:43.640
just tabbing through code yet, but I have found like

395
00:20:43.680 --> 00:20:45.640
some of these things do help. You know, there's like

396
00:20:45.680 --> 00:20:48.759
an AI assistant debugging my code too. I can remember

397
00:20:48.799 --> 00:20:52.559
again like getting started in the mainframe days, like you know,

398
00:20:52.640 --> 00:20:55.519
you had the manual, you had your best attempt, and

399
00:20:55.559 --> 00:20:57.319
you had to debug it yourself. There was no like

400
00:20:57.359 --> 00:21:00.440
going out in Google or looking on for RUMs. And

401
00:21:00.519 --> 00:21:03.839
now to today where if I've got something formatted weird

402
00:21:03.920 --> 00:21:07.200
or written slightly and correctly, you know, I can get

403
00:21:07.240 --> 00:21:09.559
help really on demand if it doesn't run and get that.

404
00:21:10.240 --> 00:21:14.319
So it has really democratized, really the ability for folks

405
00:21:14.359 --> 00:21:20.400
without that deep training to maybe do more deep analysis

406
00:21:20.480 --> 00:21:25.480
or interesting things that previously unable to. Are you seeing

407
00:21:25.519 --> 00:21:30.359
because of that our data is data becoming more consolidated

408
00:21:30.599 --> 00:21:32.920
or are there still silos that exist between you know,

409
00:21:32.960 --> 00:21:35.400
like credit and marketing and like the use case kind

410
00:21:35.400 --> 00:21:39.759
of dictates granularity and the types of data or you know,

411
00:21:39.759 --> 00:21:41.200
how do you see those things?

412
00:21:41.279 --> 00:21:41.519
Kind of?

413
00:21:41.960 --> 00:21:45.720
So, yes, there are still definitely silos of data that exist,

414
00:21:45.880 --> 00:21:49.480
and so even within a company like a bank. There's

415
00:21:50.519 --> 00:21:53.799
historically been silos right across the different subject areas, and

416
00:21:53.880 --> 00:21:56.119
the advent of the warehouse or even a cloud based

417
00:21:56.240 --> 00:22:00.240
architecture is allowing us to break down those barriers and

418
00:22:00.319 --> 00:22:04.519
build out environments where it's less out, you know, looking

419
00:22:04.599 --> 00:22:09.119
for data in the stata availability instantaneously. I think the

420
00:22:09.160 --> 00:22:11.960
other thing is this is again a little bit of

421
00:22:11.960 --> 00:22:16.039
an older buzzword, but that API architecture right, so everything

422
00:22:16.079 --> 00:22:19.359
should be published and available via API. I think the

423
00:22:19.440 --> 00:22:23.000
biggest bridge, or the biggest silo historically that is starting

424
00:22:23.000 --> 00:22:25.480
to get broken down is we take all those files

425
00:22:25.799 --> 00:22:28.759
as marketing analytics files we talked about are different feeds

426
00:22:28.759 --> 00:22:31.319
from different systems of record, and we build out this

427
00:22:31.480 --> 00:22:35.200
awesome analytical environment and we you know, we put tools

428
00:22:35.240 --> 00:22:38.640
on it, and we have time series and we can analyze, analyze, analyze.

429
00:22:38.759 --> 00:22:42.119
But the rubber hits the road when what you find

430
00:22:42.200 --> 00:22:44.799
in your analytics data you can pump over into your

431
00:22:44.839 --> 00:22:48.799
operational space and actually impact what your member or your

432
00:22:48.839 --> 00:22:53.000
customer is experiencing. And that bridge analytical to operational to

433
00:22:53.079 --> 00:22:55.839
me is the biggest silo of all and one that

434
00:22:55.880 --> 00:22:59.200
we all need to keep thinking about. How can we

435
00:22:59.279 --> 00:23:05.759
build bridges analytical to operational to make new data widgets, algorithms, alerts,

436
00:23:06.319 --> 00:23:09.920
et cetera rapidly available. You know, you can just publish

437
00:23:09.960 --> 00:23:14.640
it via API to be consumed into the operational optimization engines,

438
00:23:14.720 --> 00:23:18.839
next best engines, those sorts of things. And so, yes,

439
00:23:19.480 --> 00:23:23.240
silos definitely do still exist, and we are silo busters

440
00:23:23.319 --> 00:23:25.759
right and in that data analytics space and data scientists

441
00:23:25.759 --> 00:23:28.559
space and need to continue to do so. I think

442
00:23:28.599 --> 00:23:32.519
when you start looking across industry there are opportunities there

443
00:23:32.599 --> 00:23:37.119
as well, because there are very very large silos across

444
00:23:37.240 --> 00:23:40.960
industries that probably need to be attacked or at least

445
00:23:41.039 --> 00:23:43.200
rationalized at some point in the future to think about

446
00:23:43.200 --> 00:23:46.480
what kind of experiences and value can we bring to consumer.

447
00:23:47.400 --> 00:23:48.000
Yeah, that's good.

448
00:23:48.039 --> 00:23:50.920
I never thought of analytics, you know, being a silo

449
00:23:51.079 --> 00:23:54.400
buster type of a job. I'm sure there's someone out

450
00:23:54.480 --> 00:23:55.240
on LinkedIn with.

451
00:23:55.160 --> 00:23:56.880
That in their headline right now.

452
00:23:56.920 --> 00:24:02.000
Maybe we had that let's get a patent on that

453
00:24:02.039 --> 00:24:03.319
real quick, exactly.

454
00:24:04.200 --> 00:24:06.000
But you know you've mentioned a cloud too.

455
00:24:06.039 --> 00:24:10.119
I'm assuming it's not your own cloud that USA develops.

456
00:24:10.119 --> 00:24:11.720
So it makes me think and this has come up

457
00:24:11.720 --> 00:24:13.319
with some other guests as well, like the idea of

458
00:24:14.039 --> 00:24:17.680
bill buy or partner like as a as you're in

459
00:24:17.720 --> 00:24:20.279
this space and it's constantly changing and growing, Like, how

460
00:24:20.319 --> 00:24:23.759
do you think about this is a core competency USAA

461
00:24:23.759 --> 00:24:26.960
should own. We got to develop this in house versus uh,

462
00:24:27.039 --> 00:24:28.559
you know there are people out there that just do

463
00:24:28.599 --> 00:24:31.720
it better cheaper than us, and and let's leverage their

464
00:24:31.759 --> 00:24:34.960
their technology. Is there like a guiding principle you use

465
00:24:35.039 --> 00:24:37.319
or how do you evaluate those situations?

466
00:24:37.640 --> 00:24:40.359
Yeah, I think you have to be open to kind

467
00:24:40.400 --> 00:24:42.359
of all of the above, right, And I think it

468
00:24:42.400 --> 00:24:46.799
also goes by you know, the amount of maybe revenue

469
00:24:46.839 --> 00:24:49.319
your company has with the economies of scale you've achieved,

470
00:24:49.319 --> 00:24:53.720
do you have enough dollars to invest into yourself finding

471
00:24:53.759 --> 00:24:57.039
these new technologies or do you need help? And so

472
00:24:57.440 --> 00:25:00.200
we think of it here as at least kind My

473
00:25:00.279 --> 00:25:03.440
strategy has always been for the knowledge workers, the people

474
00:25:03.440 --> 00:25:06.440
that really understand the data and the algorithms and how

475
00:25:06.480 --> 00:25:10.720
to draw value from our particular population. People that really

476
00:25:10.720 --> 00:25:14.079
want to draw value from our particular membership, those related

477
00:25:14.119 --> 00:25:18.960
to the military or their families. I want to keep

478
00:25:19.000 --> 00:25:23.160
that knowledge on site. I want to have them be employees.

479
00:25:24.200 --> 00:25:30.319
Now if we're talking though, architecture or platform or warehouse environment.

480
00:25:31.359 --> 00:25:34.240
I find that that can be a little bit more outsourced, right,

481
00:25:34.279 --> 00:25:37.720
And so that's where we have been able to leverage,

482
00:25:38.119 --> 00:25:41.480
to your point, off prem public cloud and we have

483
00:25:41.599 --> 00:25:46.799
built pipes, large pipes. We've enhanced those pipes with AI

484
00:25:46.880 --> 00:25:49.480
and mL to test data that's going through those pipes

485
00:25:49.519 --> 00:25:53.240
out to the cloud, tested for sensitive data or for drift,

486
00:25:53.440 --> 00:25:57.319
or for anomalies, if values of different metrics or things

487
00:25:57.359 --> 00:26:03.240
are migrating beyond and acceptable range. But yeah, it really

488
00:26:03.319 --> 00:26:06.079
is that opportunity that the people that have built the

489
00:26:06.119 --> 00:26:08.440
clouds and built the tools to get the data to

490
00:26:08.480 --> 00:26:10.839
the clouds are way better at it than we are.

491
00:26:11.519 --> 00:26:14.880
That's kind of their single focus. They optimize everything. And

492
00:26:14.960 --> 00:26:18.359
so taking advantage of that was a pretty easy, easy decision,

493
00:26:18.400 --> 00:26:22.160
I believe for USAA. And then as far as like

494
00:26:22.359 --> 00:26:26.319
tools and solutions, I think that kind of goes back

495
00:26:26.359 --> 00:26:29.240
to the if you're big enough to generate and can

496
00:26:29.319 --> 00:26:33.279
afford hundreds of millions of dollars of investment into you know,

497
00:26:33.359 --> 00:26:35.720
building out your own solution. You know, there are a

498
00:26:35.759 --> 00:26:37.720
few banks that can do that. There's a few financial

499
00:26:37.720 --> 00:26:39.960
services companies that can do that. For the next tier

500
00:26:40.319 --> 00:26:44.440
it probably is one of those consortiums or finding a

501
00:26:44.519 --> 00:26:47.519
company that does it and buying them or using their products.

502
00:26:48.039 --> 00:26:50.519
But if you think about some of those problems that

503
00:26:50.559 --> 00:26:53.160
it would make sense for a consortium to tackle, like

504
00:26:53.440 --> 00:26:57.720
AML or fraud detection where we're all fighting a common enemy,

505
00:26:58.480 --> 00:27:00.559
those kinds of things make a lot of sense. You know,

506
00:27:00.599 --> 00:27:03.960
how to target matt for our product versus someone else's.

507
00:27:03.960 --> 00:27:07.480
We're not going to probably share notes on that or collaborate,

508
00:27:07.839 --> 00:27:09.920
but some of the problems I think do make sense

509
00:27:09.960 --> 00:27:14.440
for the consortium approach to get after advancing your capabilities

510
00:27:14.559 --> 00:27:15.880
using these solutions.

511
00:27:16.400 --> 00:27:18.480
Yeah, I think those are great examples, Chris, And it's

512
00:27:18.480 --> 00:27:21.759
been fun, like maybe thinking backwards a little bit to

513
00:27:21.880 --> 00:27:23.839
the things that have changed, but as you're looking to

514
00:27:23.920 --> 00:27:26.119
the future, like what are the things you've got your

515
00:27:26.160 --> 00:27:30.039
eye on for, you know, potential innovation that's continuing to happen,

516
00:27:30.119 --> 00:27:33.079
things that maybe aren't quite at a commercial scale yet,

517
00:27:33.119 --> 00:27:35.119
but you know you could see the potential like what's

518
00:27:35.160 --> 00:27:36.720
what's the future of data analytics?

519
00:27:36.759 --> 00:27:37.519
Got you're excited?

520
00:27:38.440 --> 00:27:41.079
Yeah, thanks. So there's a couple of things that we've

521
00:27:41.079 --> 00:27:44.519
been playing around with again in the ai mL space.

522
00:27:45.319 --> 00:27:49.359
Think about edge computing, So how do we put algorithms

523
00:27:49.400 --> 00:27:54.799
and identification and predictive capabilities out onto the edges of

524
00:27:54.799 --> 00:27:57.240
this wide network, whether that be on you know, your laptop,

525
00:27:57.319 --> 00:28:01.880
your iPad, your iPhone, or yourself phone of any manufacturer

526
00:28:01.920 --> 00:28:05.359
operating system. But you know, how do we put things

527
00:28:05.400 --> 00:28:09.599
there to observe and help our members, you know, figure

528
00:28:09.599 --> 00:28:12.599
out how to navigate our website or how to navigate

529
00:28:12.640 --> 00:28:16.960
our app So there's some some ideas and opportunities we

530
00:28:16.960 --> 00:28:19.279
think we have to play in that space and then

531
00:28:19.359 --> 00:28:24.119
connect that back into our systems and applications as the

532
00:28:24.279 --> 00:28:28.240
member is navigating through our processes. I think, like, if

533
00:28:28.279 --> 00:28:31.680
you think about what gets me most excited, especially working

534
00:28:31.720 --> 00:28:34.720
for USAA and for our membership, think about a scenario

535
00:28:34.759 --> 00:28:38.720
we're let's say we're we have all the data from

536
00:28:38.799 --> 00:28:41.799
Matt and his entire all of his transactions and all

537
00:28:41.799 --> 00:28:44.000
of his accounts, and all of a sudden, we see

538
00:28:44.039 --> 00:28:49.640
a sixty thousand dollars remote deposit capture from his cell

539
00:28:49.680 --> 00:28:53.920
phone of a check and then we have probably I

540
00:28:53.960 --> 00:28:57.440
don't know on average, four seconds before Matt logs off,

541
00:28:58.480 --> 00:29:01.759
and we miss the op opportunity to say something to

542
00:29:01.880 --> 00:29:05.440
Matt about that check so what is that check? Is

543
00:29:05.480 --> 00:29:09.720
it normal for Matt to deposit sixty thousand dollars or

544
00:29:09.799 --> 00:29:13.960
is it an anomaly outlier? Does it happen quarterly, annually,

545
00:29:14.279 --> 00:29:18.000
it's never happened before. All of those answers to this

546
00:29:18.160 --> 00:29:21.559
questions would help us give Matt advice on what he

547
00:29:21.599 --> 00:29:24.039
could do with that money, right, and we think that

548
00:29:24.200 --> 00:29:26.599
advice is one of our best products, and we want

549
00:29:26.640 --> 00:29:28.400
it to be one of our best products given the

550
00:29:28.400 --> 00:29:31.000
military that we serve. And so if you think about

551
00:29:31.000 --> 00:29:33.480
the data needed, So what I just described was transaction

552
00:29:33.680 --> 00:29:37.279
time series data that's been analyzed and categorized and bucketed

553
00:29:37.279 --> 00:29:42.039
into chunks where we can say is this an anomaly

554
00:29:42.160 --> 00:29:46.079
or not. But then also it's like, well it's quarterly.

555
00:29:46.160 --> 00:29:49.880
Maybe it's an investment or dividends from you know, coming

556
00:29:49.880 --> 00:29:53.440
from stocks, it's a one time thing. Maybe he just

557
00:29:53.559 --> 00:29:58.240
got a settlement for a lawsuit. Or maybe it is

558
00:30:00.799 --> 00:30:03.960
severance package. Sorry, maybe it's a severance, right, and there's

559
00:30:03.960 --> 00:30:06.880
a large and so there's a job change coming. Maybe

560
00:30:07.000 --> 00:30:11.039
this is an inheritance, and so there's again a lot

561
00:30:11.119 --> 00:30:13.200
of different things. Every one of those things I've described

562
00:30:13.240 --> 00:30:16.920
would require a very different response in those four seconds

563
00:30:17.359 --> 00:30:19.359
to talk to Matt about what he could do about it.

564
00:30:19.440 --> 00:30:22.200
Some of the options like does he have a cardlan

565
00:30:22.240 --> 00:30:24.839
that he could pay off? Does he have an investment property,

566
00:30:25.279 --> 00:30:28.519
does he have an investment portfolio that he's saving for

567
00:30:28.640 --> 00:30:33.000
in the future. So, again, all of those data elements

568
00:30:33.079 --> 00:30:37.440
coming together bring you more opportunity to say the right

569
00:30:37.480 --> 00:30:40.200
thing in those four seconds, the most meaningful thing to

570
00:30:40.279 --> 00:30:43.720
Matt giving him the very best advice I think of,

571
00:30:44.400 --> 00:30:48.519
especially again serving the military. If we've got say a

572
00:30:48.640 --> 00:30:52.240
Navy nuke, so someone that is in the Navy, they

573
00:30:52.279 --> 00:30:56.599
are becoming a nuclear specialist and they re enlist. The

574
00:30:57.160 --> 00:31:00.359
reenlistment bonus is quite sizable, and it's a very very

575
00:31:00.359 --> 00:31:04.599
important role. And so if let's say Matt was Hey,

576
00:31:04.599 --> 00:31:08.319
I just re enlisted. This was my reenlistment bonus, that

577
00:31:08.400 --> 00:31:11.759
next conversation should be, Hey, Matt, we know you're probably

578
00:31:11.799 --> 00:31:16.359
gonna go out and you buy a new Forward Lightning

579
00:31:16.599 --> 00:31:20.519
or some big, you know, expensive car. But what if

580
00:31:20.920 --> 00:31:23.039
you bought a normal car and put half of that

581
00:31:23.079 --> 00:31:26.240
into retirement right now at your age, that is going

582
00:31:26.319 --> 00:31:28.799
to set you up down the road. So imagine being

583
00:31:28.839 --> 00:31:31.799
able to have that kind of impact on a young

584
00:31:31.880 --> 00:31:35.400
service member's life because all the data was in the

585
00:31:35.440 --> 00:31:38.599
right place at the right time, the algorithms recognized it

586
00:31:38.839 --> 00:31:42.279
and put forth the best advice that they could in

587
00:31:42.319 --> 00:31:45.279
those four seconds. Right. So to me, that is the

588
00:31:45.400 --> 00:31:48.079
ultimate use case of are we doing this right yet?

589
00:31:49.039 --> 00:31:51.400
And do we have all of the data coming together

590
00:31:51.519 --> 00:31:53.759
in a way where we can even have a chance

591
00:31:54.240 --> 00:31:57.000
at putting the best advice out to one of our

592
00:31:57.000 --> 00:32:00.400
members within that very short window of time that we have.

593
00:32:01.200 --> 00:32:03.640
Yeah, that's really impactful, and I think you're right, Especially

594
00:32:03.680 --> 00:32:07.720
with a business and a membership like USAA, that advice

595
00:32:07.799 --> 00:32:09.799
or that part of the relationship that you play so

596
00:32:09.880 --> 00:32:12.960
critical and maybe brings it back to some of the

597
00:32:13.000 --> 00:32:16.559
foundations of banking and financial services, where you know, every

598
00:32:16.599 --> 00:32:18.640
transaction did involve a human right there.

599
00:32:18.839 --> 00:32:20.440
There was always someone involved.

600
00:32:20.119 --> 00:32:23.119
Usually someone you knew who had that personal connection and

601
00:32:24.200 --> 00:32:27.599
a different form of database so I just through interactions

602
00:32:27.640 --> 00:32:30.200
with the person and knowing their situation. So how do

603
00:32:30.240 --> 00:32:32.720
you take the transactional data and make that personal again?

604
00:32:32.759 --> 00:32:34.319
I think is really could be powerful.

605
00:32:35.440 --> 00:32:35.680
Cool.

606
00:32:35.839 --> 00:32:38.279
Well, Chris has been great catching up. I think we've

607
00:32:38.279 --> 00:32:40.240
covered a lot today. Is there anything maybe we missed

608
00:32:40.279 --> 00:32:42.519
that you want to cover before we wrap.

609
00:32:43.559 --> 00:32:46.920
No, it's great. I get excited getting together with the

610
00:32:47.079 --> 00:32:50.160
other people in the industry that have thought about these

611
00:32:50.240 --> 00:32:53.400
kinds of things. I get excited about what the future

612
00:32:53.440 --> 00:32:56.680
can be. I wonder how much of this can come

613
00:32:56.720 --> 00:33:00.880
to fruition within our career lifetimes, or how much are

614
00:33:00.880 --> 00:33:02.480
we going to have to hand off to our kids

615
00:33:02.519 --> 00:33:05.880
to you know, finish this work. I keep telling my kids, man,

616
00:33:06.000 --> 00:33:10.160
data and analytics is still the opportunity of this generation,

617
00:33:10.960 --> 00:33:14.000
and they do not listen. But I hope maybe yours will,

618
00:33:14.119 --> 00:33:16.759
or are other people that are listening to this podcast

619
00:33:17.039 --> 00:33:20.480
will have more success than I have had convincing people

620
00:33:20.519 --> 00:33:24.839
that this is still an exciting, highly evolving, highly impactful

621
00:33:25.599 --> 00:33:29.599
field that is going to continue to grow and move forward. Yeah.

622
00:33:29.640 --> 00:33:31.000
I don't know, Chris, I'm over three.

623
00:33:31.039 --> 00:33:32.960
I've got one who wants to be a lawyer, wan

624
00:33:33.000 --> 00:33:34.839
who wants to be an actress, and one who wants to.

625
00:33:34.880 --> 00:33:35.799
Be a YouTube star.

626
00:33:35.960 --> 00:33:42.440
So not really sorry following and dad's footsteps there, But yeah,

627
00:33:42.799 --> 00:33:44.039
a lot to be excited about.

628
00:33:44.279 --> 00:33:46.720
And thanks again, Chris, hope to catch your chick again soon.

629
00:33:47.200 --> 00:33:48.720
Yeah, of course, great man. Thanks.

630
00:33:49.359 --> 00:33:51.880
I'll start partners with banks and credit unions to grow

631
00:33:51.920 --> 00:33:55.880
households and expand consumer lending. Through its leading AI lending platform,

632
00:33:56.599 --> 00:33:59.599
up Start powered banks and credit unions leverage AI to

633
00:33:59.680 --> 00:34:03.000
offer higher approval rates and experience lower loss rates while

634
00:34:03.039 --> 00:34:07.839
simultaneously delivering the exceptional digital first lending experience that consumers demand.

635
00:34:08.360 --> 00:34:11.159
Whether you're looking to grow and enhance your existing personal

636
00:34:11.159 --> 00:34:15.000
and auto lending programs or you're just getting started, Upstart

637
00:34:15.000 --> 00:34:18.280
can help. Upstart offers an end to end solution that

638
00:34:18.320 --> 00:34:21.199
can help you find more credit worthy borrowers within your

639
00:34:21.280 --> 00:34:26.639
risk profile. With all digital underwriting, verification, loan closing, and servicing.

640
00:34:27.079 --> 00:34:30.039
It's all possible with Upstart in your corner. Learn more

641
00:34:30.079 --> 00:34:33.920
about finding new borrowers, enhancing your credit decisioning process, and

642
00:34:33.960 --> 00:34:37.719
growing your business by visiting upstart dot com slash lenders.

643
00:34:38.119 --> 00:34:40.800
That's upstart dot com slash lenders