WEBVTT
Kind: captions
Language: en
00:00:02.159 --> 00:00:05.430
Hello, I'm Josh Hollander of Horton
00:00:05.440 --> 00:00:07.590
International. I'm pleased today to be
00:00:07.600 --> 00:00:10.790
hosting another episode of the insure
00:00:10.800 --> 00:00:13.430
techch leadership podcast. I'm joined by
00:00:13.440 --> 00:00:16.470
Aaron Wright of Ernext. Aaron is a
00:00:16.480 --> 00:00:19.109
fellow of the casualty actual society
00:00:19.119 --> 00:00:22.070
who spent 20 years at USAA in pricing,
00:00:22.080 --> 00:00:23.910
underwriting, state management,
00:00:23.920 --> 00:00:25.509
catastrophe management, and data
00:00:25.519 --> 00:00:27.670
science. Then went to the other side of
00:00:27.680 --> 00:00:29.750
the table at Earnex. He now leads
00:00:29.760 --> 00:00:31.269
professional services across North
00:00:31.279 --> 00:00:32.950
America, which means he is the person
00:00:32.960 --> 00:00:35.590
who shows up after a carrier has bought
00:00:35.600 --> 00:00:37.190
the technology and has to make the
00:00:37.200 --> 00:00:39.910
transformation actually work.
00:00:39.920 --> 00:00:42.150
This conversation matters right now
00:00:42.160 --> 00:00:45.750
because Ernest has recently
00:00:45.760 --> 00:00:49.350
uh developed some research that says AI
00:00:49.360 --> 00:00:51.830
is embedded in most or some workflows at
00:00:51.840 --> 00:00:54.869
81% of insurers and that 92% of them run
00:00:54.879 --> 00:00:57.750
formal AI governance reviews, but fewer
00:00:57.760 --> 00:00:59.510
than one in three executives strongly
00:00:59.520 --> 00:01:01.270
agree that those reviews are keeping up
00:01:01.280 --> 00:01:03.670
with regulation.
00:01:03.680 --> 00:01:05.590
Welcome, Aaron.
00:01:05.600 --> 00:01:06.870
>> Thank you very much. Thanks for having
00:01:06.880 --> 00:01:09.830
me on.
00:01:09.840 --> 00:01:13.990
Yeah, absolutely. Um, you you know, uh,
00:01:14.000 --> 00:01:16.390
ve very interested to get your your take
00:01:16.400 --> 00:01:19.270
on some of these points here. Uh, before
00:01:19.280 --> 00:01:21.510
we kind of jump into things in a little
00:01:21.520 --> 00:01:23.510
bit more detail, is there anything you'd
00:01:23.520 --> 00:01:26.070
like to add beyond what I mentioned in
00:01:26.080 --> 00:01:29.190
terms of background in in in terms of
00:01:29.200 --> 00:01:31.749
um, you know, your your own uh, personal
00:01:31.759 --> 00:01:34.149
career path andor
00:01:34.159 --> 00:01:37.190
um, what you're doing today at Ernex?
00:01:37.200 --> 00:01:38.149
>> Yeah.
00:01:38.159 --> 00:01:41.590
Good question. So I I will say I I feel
00:01:41.600 --> 00:01:45.030
extreme extremely blessed uh to be where
00:01:45.040 --> 00:01:47.670
I am today and have had the career path
00:01:47.680 --> 00:01:50.230
I've had. I started off at a school with
00:01:50.240 --> 00:01:53.350
a you know a statistics undergraduate
00:01:53.360 --> 00:01:56.310
going into insurance knowing about what
00:01:56.320 --> 00:01:58.069
other people know about insurance which
00:01:58.079 --> 00:02:01.350
is I know I need to buy it. and uh
00:02:01.360 --> 00:02:03.429
quickly I mean I I started working at
00:02:03.439 --> 00:02:06.149
farmers insurance and as an actuarial
00:02:06.159 --> 00:02:08.150
analyst and I remember calling my wife
00:02:08.160 --> 00:02:11.589
after about 6 months on the job saying
00:02:11.599 --> 00:02:13.270
I'm only 6 months out of college I
00:02:13.280 --> 00:02:16.630
should be getting people coffee yet I
00:02:16.640 --> 00:02:19.589
was put in a position where I got to be
00:02:19.599 --> 00:02:20.869
on the phone with a department of
00:02:20.879 --> 00:02:24.070
insurance defending the mathematics
00:02:24.080 --> 00:02:26.949
behind a rate change and uh and I
00:02:26.959 --> 00:02:29.430
thought this is cool this is what I need
00:02:29.440 --> 00:02:32.309
to do throughout my career gotten the
00:02:32.319 --> 00:02:34.470
opportunity to uh to work not only in
00:02:34.480 --> 00:02:37.190
the pricing and actuarial space but in
00:02:37.200 --> 00:02:39.589
reinsurance and catastrophe management,
00:02:39.599 --> 00:02:41.830
state management, uh data science and
00:02:41.840 --> 00:02:45.030
predictive modeling. Uh and through all
00:02:45.040 --> 00:02:47.270
of that, you get to see how all the
00:02:47.280 --> 00:02:49.190
different pieces and all the different
00:02:49.200 --> 00:02:51.030
operational decisions really do come
00:02:51.040 --> 00:02:53.110
together. And it's important to
00:02:53.120 --> 00:02:55.750
understand across the board when you
00:02:55.760 --> 00:02:57.990
make a decision how it impacts
00:02:58.000 --> 00:03:00.869
everything. I'll also say when it comes
00:03:00.879 --> 00:03:03.990
to moving to earnex, I had not planned
00:03:04.000 --> 00:03:07.589
on leaving insurance. I had always
00:03:07.599 --> 00:03:09.670
thought that I would die an insurance
00:03:09.680 --> 00:03:13.910
executive. Uh but uh you know one thing
00:03:13.920 --> 00:03:16.309
I'll say is living on the other side of
00:03:16.319 --> 00:03:19.990
the street uh I see how much uh
00:03:20.000 --> 00:03:22.630
different companies are stuck in
00:03:22.640 --> 00:03:25.190
different ways that their technology has
00:03:25.200 --> 00:03:27.830
been holding them back. You've probably
00:03:27.840 --> 00:03:30.070
heard tale of like, you know, those
00:03:30.080 --> 00:03:31.670
experiments they've done with mice or
00:03:31.680 --> 00:03:34.470
rats where they've put uh, you know,
00:03:34.480 --> 00:03:36.550
them in mazes and said, you know, here's
00:03:36.560 --> 00:03:38.229
how you go find cheese and they do it
00:03:38.239 --> 00:03:40.390
enough times that they can take away the
00:03:40.400 --> 00:03:41.910
constraints. They can take away the
00:03:41.920 --> 00:03:43.990
other things that would prevent them
00:03:44.000 --> 00:03:45.750
from having that straight shot and
00:03:45.760 --> 00:03:47.589
they'll still do it because that's the
00:03:47.599 --> 00:03:50.710
way they're trained. I see more and more
00:03:50.720 --> 00:03:53.750
that the legacy technology the the ways
00:03:53.760 --> 00:03:56.070
that we've built to get around the
00:03:56.080 --> 00:03:58.070
problems to still you know be innovative
00:03:58.080 --> 00:04:01.910
people and do cool things uh those the
00:04:01.920 --> 00:04:04.390
technology we've had has forced us to go
00:04:04.400 --> 00:04:07.670
in certain paths and uh we really need
00:04:07.680 --> 00:04:10.869
to to get out of some of those uh I'll
00:04:10.879 --> 00:04:13.429
say mazes that we've put ourselves in in
00:04:13.439 --> 00:04:15.190
the past.
00:04:15.200 --> 00:04:17.110
You know, one thing on the on the
00:04:17.120 --> 00:04:18.870
statistics that I quoted in the in the
00:04:18.880 --> 00:04:21.509
beginning from your research is
00:04:21.519 --> 00:04:23.909
um
00:04:23.919 --> 00:04:25.830
you you know I guess in particular the
00:04:25.840 --> 00:04:28.550
second one the 92% but but really for
00:04:28.560 --> 00:04:31.830
both of them you know what does it
00:04:31.840 --> 00:04:35.510
really mean you know to say 81% have
00:04:35.520 --> 00:04:38.390
have AI embedded in most or some
00:04:38.400 --> 00:04:41.030
workflows like you know is that
00:04:41.040 --> 00:04:44.070
sophisticated is it basic and then I I
00:04:44.080 --> 00:04:47.110
don't know I I ort of find it surprising
00:04:47.120 --> 00:04:49.909
that 922%
00:04:49.919 --> 00:04:52.950
actually have AI governance reviews but
00:04:52.960 --> 00:04:55.110
again it's sort of like what's the rigor
00:04:55.120 --> 00:04:58.629
of that um etc. So you know I guess
00:04:58.639 --> 00:05:00.390
that's really the heart heart of the
00:05:00.400 --> 00:05:02.950
discussion today but
00:05:02.960 --> 00:05:05.350
>> you know maybe I'm may maybe I'm jumping
00:05:05.360 --> 00:05:08.150
too quick to the punchline but but I
00:05:08.160 --> 00:05:09.990
mean I'm really curious about those
00:05:10.000 --> 00:05:11.118
numbers.
00:05:11.128 --> 00:05:11.189
>> [snorts]
00:05:11.199 --> 00:05:13.749
>> Yeah, absolutely. So,
00:05:13.759 --> 00:05:17.909
it doesn't surprise me that uh so many
00:05:17.919 --> 00:05:20.310
people, so many executives at insurance
00:05:20.320 --> 00:05:23.590
companies say, "Oh, yes, we have AI."
00:05:23.600 --> 00:05:25.430
Right?
00:05:25.440 --> 00:05:28.790
Saying you have AI and actually having
00:05:28.800 --> 00:05:31.670
it be something that hits, you know, the
00:05:31.680 --> 00:05:34.469
balance sheet that your CFO cares about
00:05:34.479 --> 00:05:37.510
is a very different thing.
00:05:37.520 --> 00:05:40.070
there there are areas in which you know
00:05:40.080 --> 00:05:43.510
AI I think is truly making a difference
00:05:43.520 --> 00:05:45.510
and there's other areas where I think
00:05:45.520 --> 00:05:49.350
it's still kind of hype
00:05:49.360 --> 00:05:52.150
>> and and I could dig into those more but
00:05:52.160 --> 00:05:55.029
but to answer the the second part of uh
00:05:55.039 --> 00:05:57.590
the question you asked there are a lot
00:05:57.600 --> 00:05:59.270
of companies that have governance
00:05:59.280 --> 00:06:01.110
reviews right that are looking to see
00:06:01.120 --> 00:06:03.909
what the regulatory space looks like out
00:06:03.919 --> 00:06:06.710
there we're a a really conservative
00:06:06.720 --> 00:06:09.990
industry because we have to be right
00:06:10.000 --> 00:06:12.150
people are counting on us to fulfill the
00:06:12.160 --> 00:06:15.430
promise and so we have to be uh that
00:06:15.440 --> 00:06:17.830
cautious. So we're used to doing those
00:06:17.840 --> 00:06:19.909
types of things.
00:06:19.919 --> 00:06:23.110
Uh I I actually had a conversation with
00:06:23.120 --> 00:06:24.950
a senior executive from one of our
00:06:24.960 --> 00:06:27.350
customers where we were just talking
00:06:27.360 --> 00:06:29.350
about the future talking about how AI
00:06:29.360 --> 00:06:32.230
could fit into various pricing or
00:06:32.240 --> 00:06:34.390
underwriting workflows.
00:06:34.400 --> 00:06:35.990
and
00:06:36.000 --> 00:06:38.870
he made a comment something to the
00:06:38.880 --> 00:06:42.469
effect of, "Yeah, we want to do that,
00:06:42.479 --> 00:06:44.230
but it's kind of on the back burner
00:06:44.240 --> 00:06:45.749
because we want to see what regulation
00:06:45.759 --> 00:06:49.510
does." And so, while there are lots of
00:06:49.520 --> 00:06:51.350
insurance companies that have reviews,
00:06:51.360 --> 00:06:54.469
what 92% I think you said, have reviews
00:06:54.479 --> 00:06:56.629
of what's out there,
00:06:56.639 --> 00:06:58.629
less than a third are comfortable that
00:06:58.639 --> 00:07:00.150
they actually know where this is all
00:07:00.160 --> 00:07:01.670
going to go and what are the regulations
00:07:01.680 --> 00:07:04.070
going to come from. whether it's is this
00:07:04.080 --> 00:07:05.909
state going to put something new on me?
00:07:05.919 --> 00:07:07.270
Is it going to be different across
00:07:07.280 --> 00:07:08.790
states? Is the federal government going
00:07:08.800 --> 00:07:11.749
to come in and do something? Is Europe
00:07:11.759 --> 00:07:13.589
going to put stricter uh requirements
00:07:13.599 --> 00:07:14.950
out there that are going to be adopted
00:07:14.960 --> 00:07:17.909
across the board? Um
00:07:17.919 --> 00:07:20.150
there's a lot of uh I'd say just
00:07:20.160 --> 00:07:21.909
uncertainty.
00:07:21.919 --> 00:07:24.710
And so when it comes to uh you know the
00:07:24.720 --> 00:07:26.790
actual operational decisions that are
00:07:26.800 --> 00:07:30.710
making a difference that's where
00:07:30.720 --> 00:07:33.110
companies seem to be stepping into this
00:07:33.120 --> 00:07:37.510
in a very measured pace. Um now to to go
00:07:37.520 --> 00:07:39.830
back to to what you or what I was
00:07:39.840 --> 00:07:41.589
talking about earlier there are some
00:07:41.599 --> 00:07:44.870
places I think that are seeing real
00:07:44.880 --> 00:07:46.550
changes
00:07:46.560 --> 00:07:50.150
>> because of AI not just hype and I'll
00:07:50.160 --> 00:07:53.749
give you an example. So like the uh
00:07:53.759 --> 00:07:57.670
maybe the contact center um things we're
00:07:57.680 --> 00:07:59.830
actually seeing improvements in things
00:07:59.840 --> 00:08:03.430
such as you know the the cost per ticket
00:08:03.440 --> 00:08:05.189
and the handle time that's out there
00:08:05.199 --> 00:08:09.670
with claim centers um because AI is able
00:08:09.680 --> 00:08:12.070
to step in. Ernex actually has something
00:08:12.080 --> 00:08:15.110
in this space as well where we look at
00:08:15.120 --> 00:08:17.909
vertical AI really trying to give it an
00:08:17.919 --> 00:08:20.550
AI the right context, give it the right
00:08:20.560 --> 00:08:22.150
information, the right training, the
00:08:22.160 --> 00:08:24.230
right boundaries, the right governance
00:08:24.240 --> 00:08:27.510
so that for something as simple as a
00:08:27.520 --> 00:08:30.070
policyholder calling in to say why did
00:08:30.080 --> 00:08:33.750
my premium change that a call center
00:08:33.760 --> 00:08:35.990
representative can have served up to
00:08:36.000 --> 00:08:38.709
them the right information in the right
00:08:38.719 --> 00:08:42.149
context applicable to that policyholder
00:08:42.159 --> 00:08:45.590
and give a consistent answer that makes
00:08:45.600 --> 00:08:49.190
sense and is right. I've uh I've led
00:08:49.200 --> 00:08:51.030
pricing teams. I will tell you one of
00:08:51.040 --> 00:08:53.670
the scariest things to do is listen to
00:08:53.680 --> 00:08:56.230
phone calls uh of call center
00:08:56.240 --> 00:08:58.230
representatives trying to explain why
00:08:58.240 --> 00:09:00.710
premiums changed and what they did and
00:09:00.720 --> 00:09:02.389
then sitting there and just putting your
00:09:02.399 --> 00:09:04.630
head in your hands saying, "I know
00:09:04.640 --> 00:09:06.870
they're doing the best they can, but
00:09:06.880 --> 00:09:08.150
that's not right."
00:09:08.160 --> 00:09:10.389
>> Yeah. And so, you know, to get that
00:09:10.399 --> 00:09:12.870
consistent across the board is a a huge
00:09:12.880 --> 00:09:15.269
lift for our call center.
00:09:15.279 --> 00:09:16.870
>> So,
00:09:16.880 --> 00:09:19.750
I'm curious, you know, obviously USAA is
00:09:19.760 --> 00:09:22.550
a really well- reggarded organization.
00:09:22.560 --> 00:09:25.190
Um, you spent a long time there. I think
00:09:25.200 --> 00:09:28.710
you've you've made some comments um
00:09:28.720 --> 00:09:31.350
about, you know, having worked in the
00:09:31.360 --> 00:09:32.870
industry for the time that you have.
00:09:32.880 --> 00:09:36.550
You've seen things work well and then
00:09:36.560 --> 00:09:38.949
um seeing where typical inefficiencies
00:09:38.959 --> 00:09:42.630
were. Are there any things that
00:09:42.640 --> 00:09:44.389
you know you thought were were maybe
00:09:44.399 --> 00:09:46.550
unique to USAA but now being on the
00:09:46.560 --> 00:09:47.670
other side of the table you've
00:09:47.680 --> 00:09:52.150
discovered are pretty widespread?
00:09:52.160 --> 00:09:55.350
>> Yeah, absolutely.
00:09:55.360 --> 00:09:59.110
So when I was going through at USA, the
00:09:59.120 --> 00:10:00.949
last role I had there was around
00:10:00.959 --> 00:10:02.550
modernization
00:10:02.560 --> 00:10:07.110
and I was uh to put it nicely
00:10:07.120 --> 00:10:11.269
embarrassed by some of the the
00:10:11.279 --> 00:10:13.990
processes, the slowness, the technology
00:10:14.000 --> 00:10:15.829
we had in place. We had some technology
00:10:15.839 --> 00:10:18.550
that was old and we had some processes
00:10:18.560 --> 00:10:20.949
that had been built up to work around
00:10:20.959 --> 00:10:25.590
things. some that uh for example uh my
00:10:25.600 --> 00:10:28.790
pricing teams had um some of their some
00:10:28.800 --> 00:10:31.829
of my analysts were spending more time
00:10:31.839 --> 00:10:34.870
validating data handoffs than they were
00:10:34.880 --> 00:10:37.910
doing analytical work that they'd been
00:10:37.920 --> 00:10:39.509
trained to do, right? They taken all
00:10:39.519 --> 00:10:42.550
these actuarial exams to do and yet
00:10:42.560 --> 00:10:45.030
they're checking spreadsheets. Um, and
00:10:45.040 --> 00:10:48.550
so I was embarrassed and knew we needed
00:10:48.560 --> 00:10:50.630
to fix it and and I'll tell you they're
00:10:50.640 --> 00:10:52.069
they're amazing and how they're moving
00:10:52.079 --> 00:10:54.310
forward and and getting all that fixed.
00:10:54.320 --> 00:10:56.790
But when I came to Ernex and started
00:10:56.800 --> 00:10:59.269
seeing more broadly across the industry,
00:10:59.279 --> 00:11:02.069
I saw this was not a a problem with one
00:11:02.079 --> 00:11:05.350
company. This was a problem that is
00:11:05.360 --> 00:11:07.750
across the industry, right? and and it
00:11:07.760 --> 00:11:11.430
really has built up over time and it is
00:11:11.440 --> 00:11:13.829
it it's made worse
00:11:13.839 --> 00:11:16.150
as uh I'd say for some of these more
00:11:16.160 --> 00:11:18.150
innovative companies that have had
00:11:18.160 --> 00:11:20.389
legacy technology holding them back and
00:11:20.399 --> 00:11:22.470
they have built this scaffolding of
00:11:22.480 --> 00:11:24.710
workarounds around everything they have
00:11:24.720 --> 00:11:27.590
out there that almost feels like look
00:11:27.600 --> 00:11:29.750
we're not being held back by legacy
00:11:29.760 --> 00:11:32.230
technology not counting all the tech
00:11:32.240 --> 00:11:35.269
debt and the workarounds and the side
00:11:35.279 --> 00:11:37.509
projects and the ones spreadsheet that
00:11:37.519 --> 00:11:39.990
is actually the process that is holding
00:11:40.000 --> 00:11:41.829
everything up, not the system you have
00:11:41.839 --> 00:11:44.949
in place. Those are the things that I
00:11:44.959 --> 00:11:47.030
think uh sometimes hold these companies
00:11:47.040 --> 00:11:50.230
back. Um but but yes, it is I I found
00:11:50.240 --> 00:11:52.550
that it is much it is much more an
00:11:52.560 --> 00:11:55.350
industry problem for us to solve and not
00:11:55.360 --> 00:11:57.430
just a one company or another company
00:11:57.440 --> 00:11:59.910
problem to solve.
00:11:59.920 --> 00:12:01.910
>> And then one maybe from the other side.
00:12:01.920 --> 00:12:03.509
So,
00:12:03.519 --> 00:12:06.550
did did did you have sort of a a
00:12:06.560 --> 00:12:10.069
perception of I I always hate the word
00:12:10.079 --> 00:12:12.230
vendors, but for lack of a better term,
00:12:12.240 --> 00:12:15.910
vendors uh when you were inside and now
00:12:15.920 --> 00:12:17.190
that you're on the other side of the
00:12:17.200 --> 00:12:19.030
table, you know, do you see a different
00:12:19.040 --> 00:12:21.190
reality?
00:12:21.200 --> 00:12:25.590
>> Uh yes and no. Uh so I've I I think I
00:12:25.600 --> 00:12:27.990
there's times where even within earnex,
00:12:28.000 --> 00:12:31.269
I make uh my own sales team mad at me.
00:12:31.279 --> 00:12:34.150
uh when I tell them things like no no no
00:12:34.160 --> 00:12:35.750
if I got that email I wouldn't even
00:12:35.760 --> 00:12:40.310
respond. Um and I was in the position of
00:12:40.320 --> 00:12:43.829
being an insurance executive that got
00:12:43.839 --> 00:12:46.710
tens dozens of emails every day from
00:12:46.720 --> 00:12:49.350
various vendors saying we can fix all
00:12:49.360 --> 00:12:52.550
your problems. Um, the the ones that
00:12:52.560 --> 00:12:55.190
made a difference for me back then were
00:12:55.200 --> 00:12:58.150
the ones that I, you know, knew
00:12:58.160 --> 00:12:59.910
something about me, were introduced to
00:12:59.920 --> 00:13:02.470
me by someone else that, you know, I had
00:13:02.480 --> 00:13:06.629
seen actual results from, uh, that I had
00:13:06.639 --> 00:13:08.310
any sort of a personal relationship
00:13:08.320 --> 00:13:11.829
with, or that had that one tagline that
00:13:11.839 --> 00:13:13.350
that actually caught my attention
00:13:13.360 --> 00:13:14.790
because it was a problem I was already
00:13:14.800 --> 00:13:16.150
trying to fix. not something they're
00:13:16.160 --> 00:13:17.670
trying to teach me that I had a problem
00:13:17.680 --> 00:13:19.829
with, but a problem I already had that I
00:13:19.839 --> 00:13:21.990
needed to fix. When I now that I'm on
00:13:22.000 --> 00:13:25.350
the vendor side, I see how hard uh all
00:13:25.360 --> 00:13:27.509
these sales people work to really get
00:13:27.519 --> 00:13:32.470
the word out there. Um and it is uh it
00:13:32.480 --> 00:13:35.910
is sometimes difficult um you know for
00:13:35.920 --> 00:13:39.110
various executives that are pulled into
00:13:39.120 --> 00:13:41.030
the dayto-day
00:13:41.040 --> 00:13:43.910
of just keeping the business running
00:13:43.920 --> 00:13:46.230
keeping things afloat to be able to
00:13:46.240 --> 00:13:48.949
think strategically to say not just what
00:13:48.959 --> 00:13:50.710
am I doing today but what should I be
00:13:50.720 --> 00:13:52.389
doing in the future you know if if I
00:13:52.399 --> 00:13:55.110
took away the the maze that's in front
00:13:55.120 --> 00:13:57.750
of me or this you know legacy system
00:13:57.760 --> 00:13:59.910
that's forcing ing me to do one thing or
00:13:59.920 --> 00:14:01.590
another.
00:14:01.600 --> 00:14:03.350
Does that executive have enough time to
00:14:03.360 --> 00:14:05.750
really think forward thinking about what
00:14:05.760 --> 00:14:10.069
they need to do? Uh so if I had known
00:14:10.079 --> 00:14:11.829
now a little bit more what I'd known
00:14:11.839 --> 00:14:13.189
then, I think I probably would have
00:14:13.199 --> 00:14:15.269
carved out even more time to think
00:14:15.279 --> 00:14:16.389
strategically.
00:14:16.399 --> 00:14:16.790
>> Yeah.
00:14:16.800 --> 00:14:19.670
>> Back in in those positions. Uh but but
00:14:19.680 --> 00:14:21.670
that's where you know both sides have to
00:14:21.680 --> 00:14:24.389
take a grain of salt and say you know
00:14:24.399 --> 00:14:26.150
these executives and insurance companies
00:14:26.160 --> 00:14:27.910
they've got important things they got to
00:14:27.920 --> 00:14:30.870
keep running and get done and they have
00:14:30.880 --> 00:14:33.829
to think strategically. So, you know, if
00:14:33.839 --> 00:14:36.150
we on the vendor side can make sure that
00:14:36.160 --> 00:14:38.550
they they're getting what they need when
00:14:38.560 --> 00:14:40.710
they need it and being able to to be
00:14:40.720 --> 00:14:45.189
shown not just the sales pitch, but uh
00:14:45.199 --> 00:14:47.110
what is the reality and what could what
00:14:47.120 --> 00:14:48.710
could life look like in the future if
00:14:48.720 --> 00:14:51.910
they started to make some changes.
00:14:51.920 --> 00:14:54.230
So, you know, I think one of the points
00:14:54.240 --> 00:14:57.750
you've you've made is that there's
00:14:57.760 --> 00:15:00.550
um you know, technology is one thing,
00:15:00.560 --> 00:15:02.710
but there really needs to be alignment
00:15:02.720 --> 00:15:05.670
across an organization to have effective
00:15:05.680 --> 00:15:07.189
transformation.
00:15:07.199 --> 00:15:10.470
Um, someone I've I've done some some
00:15:10.480 --> 00:15:13.269
things with over the years and insure
00:15:13.279 --> 00:15:15.189
has often used the term innovation
00:15:15.199 --> 00:15:17.350
theater. So, I guess maybe I'll I'll
00:15:17.360 --> 00:15:19.990
I'll I'll sort of copy that and say, you
00:15:20.000 --> 00:15:22.790
know, what makes for real transformation
00:15:22.800 --> 00:15:24.470
versus just sort of transformation
00:15:24.480 --> 00:15:26.710
theater.
00:15:26.720 --> 00:15:30.389
>> I I like that transformation theater. Uh
00:15:30.399 --> 00:15:34.790
many sales cycles sell the demo. It's on
00:15:34.800 --> 00:15:36.550
the implementation side that we're
00:15:36.560 --> 00:15:39.030
absorbing reality that maybe didn't make
00:15:39.040 --> 00:15:43.189
it into the slide. Right. So,
00:15:43.199 --> 00:15:45.590
you talked about ownership. I think that
00:15:45.600 --> 00:15:48.470
uh I've been through quite a few
00:15:48.480 --> 00:15:51.269
implementation projects now on both
00:15:51.279 --> 00:15:55.509
sides and you're right that if ownership
00:15:55.519 --> 00:15:58.949
is not there then implementation not
00:15:58.959 --> 00:16:00.710
only goes slow but it sometimes just
00:16:00.720 --> 00:16:03.670
doesn't even happen right you it's not
00:16:03.680 --> 00:16:05.829
at implementation that you notice it
00:16:05.839 --> 00:16:09.590
it's it's 3 to 6 months into a project
00:16:09.600 --> 00:16:11.189
where
00:16:11.199 --> 00:16:13.350
somebody has to make a decision to kill
00:16:13.360 --> 00:16:15.990
a customization
00:16:16.000 --> 00:16:17.350
something that you know the company's
00:16:17.360 --> 00:16:20.310
built up over time or force a business
00:16:20.320 --> 00:16:23.829
unit to change how it works. And if
00:16:23.839 --> 00:16:26.069
there's no single owner with the real
00:16:26.079 --> 00:16:28.470
authority to actually get that done,
00:16:28.480 --> 00:16:31.749
then it's not going to happen. Most
00:16:31.759 --> 00:16:33.269
programs are not going to fail at the go
00:16:33.279 --> 00:16:34.870
live, right? They're going to fail three
00:16:34.880 --> 00:16:36.790
months before it. They're going to when
00:16:36.800 --> 00:16:39.590
somebody has to make that unpopular call
00:16:39.600 --> 00:16:41.590
and there's nobody with authority to
00:16:41.600 --> 00:16:43.990
actually make it happen. those shadow
00:16:44.000 --> 00:16:46.470
legacy processes are just going to keep
00:16:46.480 --> 00:16:50.069
going. Um, nobody's going to kill their
00:16:50.079 --> 00:16:52.629
own uh customization that they put in
00:16:52.639 --> 00:16:54.550
that they own that they they're very
00:16:54.560 --> 00:16:56.150
proud of, right? They need somebody from
00:16:56.160 --> 00:17:00.389
the top with the vision to say hold on
00:17:00.399 --> 00:17:02.470
uh we have to think about this long
00:17:02.480 --> 00:17:05.429
term, not just in small buckets or what
00:17:05.439 --> 00:17:08.150
individuals do throughout the process.
00:17:08.160 --> 00:17:11.669
The the other place where I've seen uh
00:17:11.679 --> 00:17:15.510
leadership really come into this has to
00:17:15.520 --> 00:17:20.069
do with uh parallel processes. Uh
00:17:20.079 --> 00:17:23.270
anytime you're going to put uh you know
00:17:23.280 --> 00:17:27.510
modernization into effect. Uh it's not
00:17:27.520 --> 00:17:29.110
going to be where you build it all on
00:17:29.120 --> 00:17:32.230
the side and then people come into work
00:17:32.240 --> 00:17:33.430
one day and everything will be
00:17:33.440 --> 00:17:36.549
different. Uh you're building it up.
00:17:36.559 --> 00:17:38.950
you're running parallel processes. Uh
00:17:38.960 --> 00:17:40.789
you're testing and learning into it. You
00:17:40.799 --> 00:17:42.070
maybe, you know, are doing it for one
00:17:42.080 --> 00:17:44.549
line of business or a few states within
00:17:44.559 --> 00:17:47.190
a line of business or for one team and
00:17:47.200 --> 00:17:51.270
not other teams. Um what that causes is
00:17:51.280 --> 00:17:54.710
it it's a risk to the project because
00:17:54.720 --> 00:17:56.789
you're causing real fatigue in the
00:17:56.799 --> 00:17:59.029
people that are actually saying, "You
00:17:59.039 --> 00:18:00.789
told me my job was going to get better
00:18:00.799 --> 00:18:03.669
as it was going to get easier.
00:18:03.679 --> 00:18:05.350
Yet this is not easier and this is not
00:18:05.360 --> 00:18:07.430
better. I'm running two processes at
00:18:07.440 --> 00:18:08.390
this point.
00:18:08.400 --> 00:18:08.710
>> Yeah.
00:18:08.720 --> 00:18:11.750
>> And there's burnout. So without
00:18:11.760 --> 00:18:13.350
leadership there to set the right
00:18:13.360 --> 00:18:16.230
expectations and make sure that things
00:18:16.240 --> 00:18:18.230
are progressing and to watch for that
00:18:18.240 --> 00:18:22.870
risk, then it can it can kill a project
00:18:22.880 --> 00:18:24.789
or at least kill momentum within a
00:18:24.799 --> 00:18:28.310
project.
00:18:28.320 --> 00:18:31.029
I I'm also curious given your background
00:18:31.039 --> 00:18:34.150
as an actuary, does that influence
00:18:34.160 --> 00:18:36.870
your approach in in in sort of how you
00:18:36.880 --> 00:18:39.909
advise clients and then also how you run
00:18:39.919 --> 00:18:42.710
projects?
00:18:42.720 --> 00:18:45.830
>> Oh, as an actuary, uh so I I'll tell you
00:18:45.840 --> 00:18:47.830
as as an actuary, what it does is it
00:18:47.840 --> 00:18:49.750
makes me love spreadsheets and numbers
00:18:49.760 --> 00:18:53.029
and data. Um which, uh I I say tongue
00:18:53.039 --> 00:18:55.909
and cheek, but it's true. Um I I am
00:18:55.919 --> 00:18:59.510
responsible uh you know for many many
00:18:59.520 --> 00:19:01.750
I'll say userdefined applications to
00:19:01.760 --> 00:19:04.630
work around systems in the past because
00:19:04.640 --> 00:19:06.549
I want to get things done and I've built
00:19:06.559 --> 00:19:09.669
them on my own on the side. So it it
00:19:09.679 --> 00:19:12.310
helps me I'll say within projects to
00:19:12.320 --> 00:19:15.590
sympathize with these teams that say but
00:19:15.600 --> 00:19:17.750
look at this amazing thing we've done.
00:19:17.760 --> 00:19:20.470
Uh but but as a leader, we have to be
00:19:20.480 --> 00:19:23.430
able to put some of that aside um and
00:19:23.440 --> 00:19:26.470
and be able to say where do we start?
00:19:26.480 --> 00:19:28.789
Where do we end? Where are our success
00:19:28.799 --> 00:19:31.430
metrics? Let's put some actual
00:19:31.440 --> 00:19:33.110
governance in place. Let's put some
00:19:33.120 --> 00:19:36.150
schedules, process checks to get things
00:19:36.160 --> 00:19:39.750
done. Uh so, you know, yes, being an
00:19:39.760 --> 00:19:41.510
actuary makes me love spreadsheets. It
00:19:41.520 --> 00:19:45.270
makes me love data. uh but uh growing
00:19:45.280 --> 00:19:47.909
into a leader it's helped me see how we
00:19:47.919 --> 00:19:49.430
have to have that bigger picture in
00:19:49.440 --> 00:19:50.549
mind.
00:19:50.559 --> 00:19:52.789
>> Yeah. Just one one quick question on
00:19:52.799 --> 00:19:56.950
that. So it sounds like um you know that
00:19:56.960 --> 00:19:59.909
with with all the AI tools available now
00:19:59.919 --> 00:20:02.150
that that you know the term builder is a
00:20:02.160 --> 00:20:04.310
big one. Uh you know individuals can be
00:20:04.320 --> 00:20:07.270
builders now. So, it sounds like,
00:20:07.280 --> 00:20:08.710
you know, whether it's spreadsheets or
00:20:08.720 --> 00:20:10.710
or other
00:20:10.720 --> 00:20:12.630
um sort of personal tools you developed,
00:20:12.640 --> 00:20:14.390
it sounds like you've been kind of that
00:20:14.400 --> 00:20:17.830
that that builder profile for a while
00:20:17.840 --> 00:20:21.669
now. Has the availability of of of sort
00:20:21.679 --> 00:20:23.350
of tools to use personally, not
00:20:23.360 --> 00:20:25.350
necessarily, you know, organizationwide
00:20:25.360 --> 00:20:27.350
big AI initiatives, but tools available
00:20:27.360 --> 00:20:29.590
to use personally change the way you you
00:20:29.600 --> 00:20:32.630
work or do any of that building at all?
00:20:32.640 --> 00:20:36.149
>> It does. Um, so
00:20:36.159 --> 00:20:37.750
you have to be careful when you're doing
00:20:37.760 --> 00:20:39.190
anything for work, right? Because
00:20:39.200 --> 00:20:40.470
especially in the the governed
00:20:40.480 --> 00:20:42.070
environment we live, it has to make sure
00:20:42.080 --> 00:20:43.909
that the data is protected and all the
00:20:43.919 --> 00:20:47.029
rest of that. So I will say some at work
00:20:47.039 --> 00:20:49.750
and some in my personal life, I love the
00:20:49.760 --> 00:20:52.870
availability of vibe coding, right?
00:20:52.880 --> 00:20:56.630
uh it if I have to get things done now
00:20:56.640 --> 00:20:59.510
and I want something pretty uh you know
00:20:59.520 --> 00:21:02.230
a nice way to do it with vibe coding I
00:21:02.240 --> 00:21:03.990
can usually put something together
00:21:04.000 --> 00:21:07.190
pretty darn fast. Uh if I want to use
00:21:07.200 --> 00:21:09.909
that though in an insurance environment
00:21:09.919 --> 00:21:11.350
uh that doesn't that's where it doesn't
00:21:11.360 --> 00:21:13.430
fly right if I'm building a prototype to
00:21:13.440 --> 00:21:15.909
say art of the possible that's one thing
00:21:15.919 --> 00:21:18.789
but if I want long-term governance if I
00:21:18.799 --> 00:21:21.350
want uh you know assurityity that it'll
00:21:21.360 --> 00:21:23.990
work not just for my main use case but
00:21:24.000 --> 00:21:28.549
extensible to to others uh
00:21:28.559 --> 00:21:30.549
I am happy to build a prototype with
00:21:30.559 --> 00:21:34.710
vibe coding um and show it to my IT team
00:21:34.720 --> 00:21:37.029
show it to my developers and say, "No,
00:21:37.039 --> 00:21:39.190
no, no. I'm not just going to write up a
00:21:39.200 --> 00:21:40.870
description about what I needed to do. I
00:21:40.880 --> 00:21:42.950
can show you here's what it does in the
00:21:42.960 --> 00:21:45.830
best use case." But I understand that on
00:21:45.840 --> 00:21:48.870
my side, um, that there's too much at
00:21:48.880 --> 00:21:50.870
stake within insurance companies to
00:21:50.880 --> 00:21:53.830
trust somebody like me to vibe code
00:21:53.840 --> 00:21:56.070
something up to use on a long-term
00:21:56.080 --> 00:21:57.190
basis.
00:21:57.200 --> 00:22:00.310
>> Yeah. So, I want to make sure we we get
00:22:00.320 --> 00:22:02.149
to
00:22:02.159 --> 00:22:06.149
um the agents that that uh Ernex
00:22:06.159 --> 00:22:09.430
launched uh earlier this year, I think a
00:22:09.440 --> 00:22:13.590
month or or so ago. Um could you share a
00:22:13.600 --> 00:22:15.590
little bit more about you know what you
00:22:15.600 --> 00:22:18.230
guys are are offering and you know what
00:22:18.240 --> 00:22:21.190
what you see agents actually doing in a
00:22:21.200 --> 00:22:23.510
carrier today and you know is it
00:22:23.520 --> 00:22:25.110
complementing what people are doing? Is
00:22:25.120 --> 00:22:27.830
it yeah, you know, replacing what people
00:22:27.840 --> 00:22:30.070
are doing and and how it's kind of
00:22:30.080 --> 00:22:32.580
fitting into the to the workflows.
00:22:32.590 --> 00:22:33.029
[snorts]
00:22:33.039 --> 00:22:36.549
>> Yeah, that's a great question. So, so
00:22:36.559 --> 00:22:38.950
let me start by just saying in general
00:22:38.960 --> 00:22:43.029
um our view of how AI should be used
00:22:43.039 --> 00:22:46.549
within insurance is to enable expertise,
00:22:46.559 --> 00:22:49.430
right? uh there's probably going to be
00:22:49.440 --> 00:22:53.270
some jobs that uh you know down the line
00:22:53.280 --> 00:22:55.750
could be are are going to be impacted,
00:22:55.760 --> 00:22:58.310
you know, but the reality is we're not
00:22:58.320 --> 00:22:59.830
seeing companies and we probably
00:22:59.840 --> 00:23:01.270
shouldn't be seeing companies that are
00:23:01.280 --> 00:23:02.870
saying I'm going to replace 20% of my
00:23:02.880 --> 00:23:06.710
workforce with AI. That's that's not the
00:23:06.720 --> 00:23:09.029
right stage we're at right now. where we
00:23:09.039 --> 00:23:11.830
are now is we need to be saying for the
00:23:11.840 --> 00:23:13.350
experts that are out there, how do we
00:23:13.360 --> 00:23:15.110
take some of the lowhanging fruit, the
00:23:15.120 --> 00:23:16.390
things that they're doing within their
00:23:16.400 --> 00:23:19.750
job and make it easier. Uh companies
00:23:19.760 --> 00:23:22.149
have innovation all the time that sits
00:23:22.159 --> 00:23:24.149
on the shelf because they just can't get
00:23:24.159 --> 00:23:27.990
it into production. AI is no different.
00:23:28.000 --> 00:23:32.230
So what we at Ernex have really uh are
00:23:32.240 --> 00:23:33.750
launching right now and have already
00:23:33.760 --> 00:23:35.270
launched earlier this year and starting
00:23:35.280 --> 00:23:38.230
to use some of our customers is the
00:23:38.240 --> 00:23:40.710
concept of uh AIOS. And while I won't
00:23:40.720 --> 00:23:43.830
get into the the full uh talk track
00:23:43.840 --> 00:23:46.870
here, what I will tell you is that uh we
00:23:46.880 --> 00:23:50.789
allow for verticalized contextualized AI
00:23:50.799 --> 00:23:53.270
within the earnex and uh insurance
00:23:53.280 --> 00:23:55.990
framework to be embedded in the workflow
00:23:56.000 --> 00:23:58.549
of what it is that our customers are
00:23:58.559 --> 00:24:01.750
doing today. helping them actually see
00:24:01.760 --> 00:24:04.390
um how different parts of the flow and
00:24:04.400 --> 00:24:07.590
the process actually can be governed so
00:24:07.600 --> 00:24:09.750
that when as that regulation is evolving
00:24:09.760 --> 00:24:12.070
that it's all controlled. You know
00:24:12.080 --> 00:24:13.750
exactly what's there, how it's being
00:24:13.760 --> 00:24:16.230
done, how it's being operationalized
00:24:16.240 --> 00:24:17.990
um and it's actually helping the people
00:24:18.000 --> 00:24:19.269
at the point where they're making
00:24:19.279 --> 00:24:21.750
decisions. I gave an example a little
00:24:21.760 --> 00:24:23.269
earlier as we're in our discussion
00:24:23.279 --> 00:24:24.870
around those call center
00:24:24.880 --> 00:24:27.909
representatives. Um, we had one of our
00:24:27.919 --> 00:24:31.669
customers tell us that the the tenure
00:24:31.679 --> 00:24:34.230
the average tenure for call center
00:24:34.240 --> 00:24:35.990
representatives is about 12 months at
00:24:36.000 --> 00:24:38.149
their company and it takes 18 months to
00:24:38.159 --> 00:24:40.549
get them fully trained. Well, that
00:24:40.559 --> 00:24:43.190
doesn't work. And so, if you can instead
00:24:43.200 --> 00:24:46.470
be able to embed this verticalized AI
00:24:46.480 --> 00:24:49.830
within to do something like explain uh
00:24:49.840 --> 00:24:52.470
why premium changed, right? that that
00:24:52.480 --> 00:24:55.269
makes it not only so that the training
00:24:55.279 --> 00:24:57.990
burden that's on this is shrunk so that
00:24:58.000 --> 00:25:01.350
you're getting them up and going faster,
00:25:01.360 --> 00:25:03.590
but I also wonder while I haven't seen
00:25:03.600 --> 00:25:06.310
numbers on this if it could extend that
00:25:06.320 --> 00:25:08.310
12 months to something longer by making
00:25:08.320 --> 00:25:11.269
somebody's job that much easier and that
00:25:11.279 --> 00:25:13.269
much better. Right? Not only is the
00:25:13.279 --> 00:25:15.990
company having good consistent
00:25:16.000 --> 00:25:18.070
explanations from customer to customer
00:25:18.080 --> 00:25:21.110
that match reality, uh, but for a call
00:25:21.120 --> 00:25:23.909
center agent who is juggling everything
00:25:23.919 --> 00:25:25.990
under the sun to be able to have
00:25:26.000 --> 00:25:28.070
something brought directly to them that
00:25:28.080 --> 00:25:30.470
it has access to all the operational
00:25:30.480 --> 00:25:34.070
decisions, the customer's profile, the
00:25:34.080 --> 00:25:36.070
right constraints put around it around
00:25:36.080 --> 00:25:38.310
what language should be there um, and
00:25:38.320 --> 00:25:40.149
served up to them right when they need
00:25:40.159 --> 00:25:42.710
it in the tools they're already
00:25:42.720 --> 00:25:45.510
That's gold and that's what makes uh
00:25:45.520 --> 00:25:47.350
makes things actually work and better
00:25:47.360 --> 00:25:49.590
and that's what'll flow through then to
00:25:49.600 --> 00:25:52.470
your CFO's bottom line.
00:25:52.480 --> 00:25:54.950
>> So I don't even know the right time
00:25:54.960 --> 00:25:57.269
frame to use anymore. You know in the
00:25:57.279 --> 00:25:58.789
past I would have said three years out
00:25:58.799 --> 00:26:00.950
but things are moving so quickly. I I
00:26:00.960 --> 00:26:03.590
don't know what what what to use. But
00:26:03.600 --> 00:26:09.669
you know what what is a team actually
00:26:09.679 --> 00:26:10.870
you know and maybe you can talk about
00:26:10.880 --> 00:26:13.430
individual personas but what is a team
00:26:13.440 --> 00:26:16.310
actually doing sort of all day once this
00:26:16.320 --> 00:26:18.870
is really working and and embedded in
00:26:18.880 --> 00:26:21.669
our organizations.
00:26:21.679 --> 00:26:24.950
Yeah, I I don't think that the
00:26:24.960 --> 00:26:28.470
the steps needed to price an insurance
00:26:28.480 --> 00:26:32.149
policy or to determine the right, you
00:26:32.159 --> 00:26:34.070
know, underwriting analytics and, you
00:26:34.080 --> 00:26:36.070
know, knock criteria, underwriting
00:26:36.080 --> 00:26:37.430
guidance. I don't think that the
00:26:37.440 --> 00:26:39.750
fundamentals around that changes. I
00:26:39.760 --> 00:26:42.630
think what does change is that the
00:26:42.640 --> 00:26:45.909
decisions that are being made, the speed
00:26:45.919 --> 00:26:48.870
it takes to get things done. uh for
00:26:48.880 --> 00:26:51.269
example you know a data scientist that's
00:26:51.279 --> 00:26:55.269
building a loss cost model uh for you
00:26:55.279 --> 00:26:57.750
know insurance pricing
00:26:57.760 --> 00:27:00.630
when I did some of the first ones back
00:27:00.640 --> 00:27:03.190
in the early 2000s it took me forever it
00:27:03.200 --> 00:27:05.830
took me months to get it done because I
00:27:05.840 --> 00:27:07.269
didn't know what I was doing the
00:27:07.279 --> 00:27:09.669
technology was new
00:27:09.679 --> 00:27:13.110
nowadays there's actually and I'll tell
00:27:13.120 --> 00:27:14.950
you earnex has some of this embedded in
00:27:14.960 --> 00:27:17.750
our data science platform abilities for
00:27:17.760 --> 00:27:19.750
AI to jump in and say, "Let me tell you
00:27:19.760 --> 00:27:21.590
the difference between two versions of a
00:27:21.600 --> 00:27:24.710
model. Let me suggest variables that you
00:27:24.720 --> 00:27:27.430
should consider. Let me look at the fits
00:27:27.440 --> 00:27:29.830
that you have within there or the the
00:27:29.840 --> 00:27:31.909
bucketing of your different variables,
00:27:31.919 --> 00:27:33.990
smoothing capabilities." I mean, all of
00:27:34.000 --> 00:27:35.830
these different things that a data
00:27:35.840 --> 00:27:39.350
scientist has to consider. Now, we're
00:27:39.360 --> 00:27:40.950
not taking away the need for their
00:27:40.960 --> 00:27:43.029
expertise, but what we are doing is
00:27:43.039 --> 00:27:44.710
making it so that as they're doing their
00:27:44.720 --> 00:27:46.950
job, it's smoother along the way. And I
00:27:46.960 --> 00:27:48.310
think that's across the board with all
00:27:48.320 --> 00:27:50.070
these different experts that we're
00:27:50.080 --> 00:27:53.029
looking at that, you know, the value
00:27:53.039 --> 00:27:55.750
added work that they do today will be
00:27:55.760 --> 00:27:58.389
there and will be magnified. It's just
00:27:58.399 --> 00:28:00.230
the stuff that takes away from them
00:28:00.240 --> 00:28:02.230
actually using some of that knowledge
00:28:02.240 --> 00:28:04.389
and ability they have that needs to fall
00:28:04.399 --> 00:28:07.350
by the wayside. uh taking a spreadsheet
00:28:07.360 --> 00:28:08.950
and double-checking it between, you
00:28:08.960 --> 00:28:11.110
know, two different screens or copying
00:28:11.120 --> 00:28:15.190
and pasting data from an email into an
00:28:15.200 --> 00:28:17.909
operational system. That that shouldn't
00:28:17.919 --> 00:28:19.590
have to be somebody's job. It's not
00:28:19.600 --> 00:28:22.149
something somebody enjoys doing, and it
00:28:22.159 --> 00:28:23.350
should be something that we're able to
00:28:23.360 --> 00:28:26.230
hand to an AI and make it help uh where
00:28:26.240 --> 00:28:28.470
it's supposed to.
00:28:28.480 --> 00:28:30.310
>> So, Aaron, I really appreciate your
00:28:30.320 --> 00:28:32.230
insights. Is there anything we haven't
00:28:32.240 --> 00:28:33.750
touched on today that you'd like to
00:28:33.760 --> 00:28:37.909
cover before we wrap up?
00:28:37.919 --> 00:28:41.830
>> Maybe just at the end I I'll say that uh
00:28:41.840 --> 00:28:44.630
you know I through my career I have seen
00:28:44.640 --> 00:28:49.510
where uh legacy systems have really
00:28:49.520 --> 00:28:52.070
slowed companies down and while yes
00:28:52.080 --> 00:28:53.830
there are workarounds those are still
00:28:53.840 --> 00:28:57.029
another form of slowing us down. I I
00:28:57.039 --> 00:28:59.029
think as an industry we really need to
00:28:59.039 --> 00:29:01.510
take a hard look at what the art of the
00:29:01.520 --> 00:29:03.590
possible is in the future and we're
00:29:03.600 --> 00:29:05.830
getting pressure from both sides from
00:29:05.840 --> 00:29:07.750
you know legacy systems slowing us down
00:29:07.760 --> 00:29:09.669
from new technology with AI that's out
00:29:09.679 --> 00:29:12.070
there that can allow us to speed up and
00:29:12.080 --> 00:29:14.070
where we meet in the middle is going to
00:29:14.080 --> 00:29:16.149
be how we take our industry forward and
00:29:16.159 --> 00:29:17.510
I think it's important for us to give
00:29:17.520 --> 00:29:20.549
that the right amount of thought and uh
00:29:20.559 --> 00:29:22.470
time and effort to make sure that that
00:29:22.480 --> 00:29:24.389
happens and that we as an industry can
00:29:24.399 --> 00:29:27.510
actually move forward in the right
00:29:27.520 --> 00:29:31.029
Yeah, I mean I uh I do think sometimes
00:29:31.039 --> 00:29:33.029
there's there's too much do and not
00:29:33.039 --> 00:29:35.350
enough think. um you know sort of ready
00:29:35.360 --> 00:29:38.389
fire aim ready fire aim situation
00:29:38.399 --> 00:29:42.070
because you know and I do think to your
00:29:42.080 --> 00:29:45.990
point before you know insurance it you
00:29:46.000 --> 00:29:47.430
know it's an industry that's all about
00:29:47.440 --> 00:29:49.510
risk management so it tends to be
00:29:49.520 --> 00:29:51.029
conservative
00:29:51.039 --> 00:29:53.909
but you know inevitably it does face a
00:29:53.919 --> 00:29:55.510
lot of the same pressures that that
00:29:55.520 --> 00:29:58.070
every other industry does um you know do
00:29:58.080 --> 00:30:01.190
more do faster do more efficiently do
00:30:01.200 --> 00:30:07.669
cheaper um And um and uh
00:30:07.679 --> 00:30:12.549
you know it can and it can create uh uh
00:30:12.559 --> 00:30:14.710
you know in this this environment uh
00:30:14.720 --> 00:30:17.190
sort of that that that FOMO and and and
00:30:17.200 --> 00:30:20.470
and people can send them off down a path
00:30:20.480 --> 00:30:22.789
that that doesn't always make the most
00:30:22.799 --> 00:30:24.230
sense just because they feel like they
00:30:24.240 --> 00:30:26.149
have to do something.
00:30:26.159 --> 00:30:27.990
>> Yeah. Very true.
00:30:28.000 --> 00:30:28.389
>> Yeah.
00:30:28.399 --> 00:30:30.950
>> Very true. Well, Aaron Wright, VP of
00:30:30.960 --> 00:30:33.990
Professional Services Americas at Ernex,
00:30:34.000 --> 00:30:36.070
thanks so much for joining me on this
00:30:36.080 --> 00:30:38.389
episode of the Insure Tech Leadership
00:30:38.399 --> 00:30:39.750
Podcast.
00:30:39.760 --> 00:30:43.120
>> Thank you, Josh. I love
Kind: captions
Language: en
00:00:02.159 --> 00:00:05.430
Hello, I'm Josh Hollander of Horton
00:00:05.440 --> 00:00:07.590
International. I'm pleased today to be
00:00:07.600 --> 00:00:10.790
hosting another episode of the insure
00:00:10.800 --> 00:00:13.430
techch leadership podcast. I'm joined by
00:00:13.440 --> 00:00:16.470
Aaron Wright of Ernext. Aaron is a
00:00:16.480 --> 00:00:19.109
fellow of the casualty actual society
00:00:19.119 --> 00:00:22.070
who spent 20 years at USAA in pricing,
00:00:22.080 --> 00:00:23.910
underwriting, state management,
00:00:23.920 --> 00:00:25.509
catastrophe management, and data
00:00:25.519 --> 00:00:27.670
science. Then went to the other side of
00:00:27.680 --> 00:00:29.750
the table at Earnex. He now leads
00:00:29.760 --> 00:00:31.269
professional services across North
00:00:31.279 --> 00:00:32.950
America, which means he is the person
00:00:32.960 --> 00:00:35.590
who shows up after a carrier has bought
00:00:35.600 --> 00:00:37.190
the technology and has to make the
00:00:37.200 --> 00:00:39.910
transformation actually work.
00:00:39.920 --> 00:00:42.150
This conversation matters right now
00:00:42.160 --> 00:00:45.750
because Ernest has recently
00:00:45.760 --> 00:00:49.350
uh developed some research that says AI
00:00:49.360 --> 00:00:51.830
is embedded in most or some workflows at
00:00:51.840 --> 00:00:54.869
81% of insurers and that 92% of them run
00:00:54.879 --> 00:00:57.750
formal AI governance reviews, but fewer
00:00:57.760 --> 00:00:59.510
than one in three executives strongly
00:00:59.520 --> 00:01:01.270
agree that those reviews are keeping up
00:01:01.280 --> 00:01:03.670
with regulation.
00:01:03.680 --> 00:01:05.590
Welcome, Aaron.
00:01:05.600 --> 00:01:06.870
>> Thank you very much. Thanks for having
00:01:06.880 --> 00:01:09.830
me on.
00:01:09.840 --> 00:01:13.990
Yeah, absolutely. Um, you you know, uh,
00:01:14.000 --> 00:01:16.390
ve very interested to get your your take
00:01:16.400 --> 00:01:19.270
on some of these points here. Uh, before
00:01:19.280 --> 00:01:21.510
we kind of jump into things in a little
00:01:21.520 --> 00:01:23.510
bit more detail, is there anything you'd
00:01:23.520 --> 00:01:26.070
like to add beyond what I mentioned in
00:01:26.080 --> 00:01:29.190
terms of background in in in terms of
00:01:29.200 --> 00:01:31.749
um, you know, your your own uh, personal
00:01:31.759 --> 00:01:34.149
career path andor
00:01:34.159 --> 00:01:37.190
um, what you're doing today at Ernex?
00:01:37.200 --> 00:01:38.149
>> Yeah.
00:01:38.159 --> 00:01:41.590
Good question. So I I will say I I feel
00:01:41.600 --> 00:01:45.030
extreme extremely blessed uh to be where
00:01:45.040 --> 00:01:47.670
I am today and have had the career path
00:01:47.680 --> 00:01:50.230
I've had. I started off at a school with
00:01:50.240 --> 00:01:53.350
a you know a statistics undergraduate
00:01:53.360 --> 00:01:56.310
going into insurance knowing about what
00:01:56.320 --> 00:01:58.069
other people know about insurance which
00:01:58.079 --> 00:02:01.350
is I know I need to buy it. and uh
00:02:01.360 --> 00:02:03.429
quickly I mean I I started working at
00:02:03.439 --> 00:02:06.149
farmers insurance and as an actuarial
00:02:06.159 --> 00:02:08.150
analyst and I remember calling my wife
00:02:08.160 --> 00:02:11.589
after about 6 months on the job saying
00:02:11.599 --> 00:02:13.270
I'm only 6 months out of college I
00:02:13.280 --> 00:02:16.630
should be getting people coffee yet I
00:02:16.640 --> 00:02:19.589
was put in a position where I got to be
00:02:19.599 --> 00:02:20.869
on the phone with a department of
00:02:20.879 --> 00:02:24.070
insurance defending the mathematics
00:02:24.080 --> 00:02:26.949
behind a rate change and uh and I
00:02:26.959 --> 00:02:29.430
thought this is cool this is what I need
00:02:29.440 --> 00:02:32.309
to do throughout my career gotten the
00:02:32.319 --> 00:02:34.470
opportunity to uh to work not only in
00:02:34.480 --> 00:02:37.190
the pricing and actuarial space but in
00:02:37.200 --> 00:02:39.589
reinsurance and catastrophe management,
00:02:39.599 --> 00:02:41.830
state management, uh data science and
00:02:41.840 --> 00:02:45.030
predictive modeling. Uh and through all
00:02:45.040 --> 00:02:47.270
of that, you get to see how all the
00:02:47.280 --> 00:02:49.190
different pieces and all the different
00:02:49.200 --> 00:02:51.030
operational decisions really do come
00:02:51.040 --> 00:02:53.110
together. And it's important to
00:02:53.120 --> 00:02:55.750
understand across the board when you
00:02:55.760 --> 00:02:57.990
make a decision how it impacts
00:02:58.000 --> 00:03:00.869
everything. I'll also say when it comes
00:03:00.879 --> 00:03:03.990
to moving to earnex, I had not planned
00:03:04.000 --> 00:03:07.589
on leaving insurance. I had always
00:03:07.599 --> 00:03:09.670
thought that I would die an insurance
00:03:09.680 --> 00:03:13.910
executive. Uh but uh you know one thing
00:03:13.920 --> 00:03:16.309
I'll say is living on the other side of
00:03:16.319 --> 00:03:19.990
the street uh I see how much uh
00:03:20.000 --> 00:03:22.630
different companies are stuck in
00:03:22.640 --> 00:03:25.190
different ways that their technology has
00:03:25.200 --> 00:03:27.830
been holding them back. You've probably
00:03:27.840 --> 00:03:30.070
heard tale of like, you know, those
00:03:30.080 --> 00:03:31.670
experiments they've done with mice or
00:03:31.680 --> 00:03:34.470
rats where they've put uh, you know,
00:03:34.480 --> 00:03:36.550
them in mazes and said, you know, here's
00:03:36.560 --> 00:03:38.229
how you go find cheese and they do it
00:03:38.239 --> 00:03:40.390
enough times that they can take away the
00:03:40.400 --> 00:03:41.910
constraints. They can take away the
00:03:41.920 --> 00:03:43.990
other things that would prevent them
00:03:44.000 --> 00:03:45.750
from having that straight shot and
00:03:45.760 --> 00:03:47.589
they'll still do it because that's the
00:03:47.599 --> 00:03:50.710
way they're trained. I see more and more
00:03:50.720 --> 00:03:53.750
that the legacy technology the the ways
00:03:53.760 --> 00:03:56.070
that we've built to get around the
00:03:56.080 --> 00:03:58.070
problems to still you know be innovative
00:03:58.080 --> 00:04:01.910
people and do cool things uh those the
00:04:01.920 --> 00:04:04.390
technology we've had has forced us to go
00:04:04.400 --> 00:04:07.670
in certain paths and uh we really need
00:04:07.680 --> 00:04:10.869
to to get out of some of those uh I'll
00:04:10.879 --> 00:04:13.429
say mazes that we've put ourselves in in
00:04:13.439 --> 00:04:15.190
the past.
00:04:15.200 --> 00:04:17.110
You know, one thing on the on the
00:04:17.120 --> 00:04:18.870
statistics that I quoted in the in the
00:04:18.880 --> 00:04:21.509
beginning from your research is
00:04:21.519 --> 00:04:23.909
um
00:04:23.919 --> 00:04:25.830
you you know I guess in particular the
00:04:25.840 --> 00:04:28.550
second one the 92% but but really for
00:04:28.560 --> 00:04:31.830
both of them you know what does it
00:04:31.840 --> 00:04:35.510
really mean you know to say 81% have
00:04:35.520 --> 00:04:38.390
have AI embedded in most or some
00:04:38.400 --> 00:04:41.030
workflows like you know is that
00:04:41.040 --> 00:04:44.070
sophisticated is it basic and then I I
00:04:44.080 --> 00:04:47.110
don't know I I ort of find it surprising
00:04:47.120 --> 00:04:49.909
that 922%
00:04:49.919 --> 00:04:52.950
actually have AI governance reviews but
00:04:52.960 --> 00:04:55.110
again it's sort of like what's the rigor
00:04:55.120 --> 00:04:58.629
of that um etc. So you know I guess
00:04:58.639 --> 00:05:00.390
that's really the heart heart of the
00:05:00.400 --> 00:05:02.950
discussion today but
00:05:02.960 --> 00:05:05.350
>> you know maybe I'm may maybe I'm jumping
00:05:05.360 --> 00:05:08.150
too quick to the punchline but but I
00:05:08.160 --> 00:05:09.990
mean I'm really curious about those
00:05:10.000 --> 00:05:11.118
numbers.
00:05:11.128 --> 00:05:11.189
>> [snorts]
00:05:11.199 --> 00:05:13.749
>> Yeah, absolutely. So,
00:05:13.759 --> 00:05:17.909
it doesn't surprise me that uh so many
00:05:17.919 --> 00:05:20.310
people, so many executives at insurance
00:05:20.320 --> 00:05:23.590
companies say, "Oh, yes, we have AI."
00:05:23.600 --> 00:05:25.430
Right?
00:05:25.440 --> 00:05:28.790
Saying you have AI and actually having
00:05:28.800 --> 00:05:31.670
it be something that hits, you know, the
00:05:31.680 --> 00:05:34.469
balance sheet that your CFO cares about
00:05:34.479 --> 00:05:37.510
is a very different thing.
00:05:37.520 --> 00:05:40.070
there there are areas in which you know
00:05:40.080 --> 00:05:43.510
AI I think is truly making a difference
00:05:43.520 --> 00:05:45.510
and there's other areas where I think
00:05:45.520 --> 00:05:49.350
it's still kind of hype
00:05:49.360 --> 00:05:52.150
>> and and I could dig into those more but
00:05:52.160 --> 00:05:55.029
but to answer the the second part of uh
00:05:55.039 --> 00:05:57.590
the question you asked there are a lot
00:05:57.600 --> 00:05:59.270
of companies that have governance
00:05:59.280 --> 00:06:01.110
reviews right that are looking to see
00:06:01.120 --> 00:06:03.909
what the regulatory space looks like out
00:06:03.919 --> 00:06:06.710
there we're a a really conservative
00:06:06.720 --> 00:06:09.990
industry because we have to be right
00:06:10.000 --> 00:06:12.150
people are counting on us to fulfill the
00:06:12.160 --> 00:06:15.430
promise and so we have to be uh that
00:06:15.440 --> 00:06:17.830
cautious. So we're used to doing those
00:06:17.840 --> 00:06:19.909
types of things.
00:06:19.919 --> 00:06:23.110
Uh I I actually had a conversation with
00:06:23.120 --> 00:06:24.950
a senior executive from one of our
00:06:24.960 --> 00:06:27.350
customers where we were just talking
00:06:27.360 --> 00:06:29.350
about the future talking about how AI
00:06:29.360 --> 00:06:32.230
could fit into various pricing or
00:06:32.240 --> 00:06:34.390
underwriting workflows.
00:06:34.400 --> 00:06:35.990
and
00:06:36.000 --> 00:06:38.870
he made a comment something to the
00:06:38.880 --> 00:06:42.469
effect of, "Yeah, we want to do that,
00:06:42.479 --> 00:06:44.230
but it's kind of on the back burner
00:06:44.240 --> 00:06:45.749
because we want to see what regulation
00:06:45.759 --> 00:06:49.510
does." And so, while there are lots of
00:06:49.520 --> 00:06:51.350
insurance companies that have reviews,
00:06:51.360 --> 00:06:54.469
what 92% I think you said, have reviews
00:06:54.479 --> 00:06:56.629
of what's out there,
00:06:56.639 --> 00:06:58.629
less than a third are comfortable that
00:06:58.639 --> 00:07:00.150
they actually know where this is all
00:07:00.160 --> 00:07:01.670
going to go and what are the regulations
00:07:01.680 --> 00:07:04.070
going to come from. whether it's is this
00:07:04.080 --> 00:07:05.909
state going to put something new on me?
00:07:05.919 --> 00:07:07.270
Is it going to be different across
00:07:07.280 --> 00:07:08.790
states? Is the federal government going
00:07:08.800 --> 00:07:11.749
to come in and do something? Is Europe
00:07:11.759 --> 00:07:13.589
going to put stricter uh requirements
00:07:13.599 --> 00:07:14.950
out there that are going to be adopted
00:07:14.960 --> 00:07:17.909
across the board? Um
00:07:17.919 --> 00:07:20.150
there's a lot of uh I'd say just
00:07:20.160 --> 00:07:21.909
uncertainty.
00:07:21.919 --> 00:07:24.710
And so when it comes to uh you know the
00:07:24.720 --> 00:07:26.790
actual operational decisions that are
00:07:26.800 --> 00:07:30.710
making a difference that's where
00:07:30.720 --> 00:07:33.110
companies seem to be stepping into this
00:07:33.120 --> 00:07:37.510
in a very measured pace. Um now to to go
00:07:37.520 --> 00:07:39.830
back to to what you or what I was
00:07:39.840 --> 00:07:41.589
talking about earlier there are some
00:07:41.599 --> 00:07:44.870
places I think that are seeing real
00:07:44.880 --> 00:07:46.550
changes
00:07:46.560 --> 00:07:50.150
>> because of AI not just hype and I'll
00:07:50.160 --> 00:07:53.749
give you an example. So like the uh
00:07:53.759 --> 00:07:57.670
maybe the contact center um things we're
00:07:57.680 --> 00:07:59.830
actually seeing improvements in things
00:07:59.840 --> 00:08:03.430
such as you know the the cost per ticket
00:08:03.440 --> 00:08:05.189
and the handle time that's out there
00:08:05.199 --> 00:08:09.670
with claim centers um because AI is able
00:08:09.680 --> 00:08:12.070
to step in. Ernex actually has something
00:08:12.080 --> 00:08:15.110
in this space as well where we look at
00:08:15.120 --> 00:08:17.909
vertical AI really trying to give it an
00:08:17.919 --> 00:08:20.550
AI the right context, give it the right
00:08:20.560 --> 00:08:22.150
information, the right training, the
00:08:22.160 --> 00:08:24.230
right boundaries, the right governance
00:08:24.240 --> 00:08:27.510
so that for something as simple as a
00:08:27.520 --> 00:08:30.070
policyholder calling in to say why did
00:08:30.080 --> 00:08:33.750
my premium change that a call center
00:08:33.760 --> 00:08:35.990
representative can have served up to
00:08:36.000 --> 00:08:38.709
them the right information in the right
00:08:38.719 --> 00:08:42.149
context applicable to that policyholder
00:08:42.159 --> 00:08:45.590
and give a consistent answer that makes
00:08:45.600 --> 00:08:49.190
sense and is right. I've uh I've led
00:08:49.200 --> 00:08:51.030
pricing teams. I will tell you one of
00:08:51.040 --> 00:08:53.670
the scariest things to do is listen to
00:08:53.680 --> 00:08:56.230
phone calls uh of call center
00:08:56.240 --> 00:08:58.230
representatives trying to explain why
00:08:58.240 --> 00:09:00.710
premiums changed and what they did and
00:09:00.720 --> 00:09:02.389
then sitting there and just putting your
00:09:02.399 --> 00:09:04.630
head in your hands saying, "I know
00:09:04.640 --> 00:09:06.870
they're doing the best they can, but
00:09:06.880 --> 00:09:08.150
that's not right."
00:09:08.160 --> 00:09:10.389
>> Yeah. And so, you know, to get that
00:09:10.399 --> 00:09:12.870
consistent across the board is a a huge
00:09:12.880 --> 00:09:15.269
lift for our call center.
00:09:15.279 --> 00:09:16.870
>> So,
00:09:16.880 --> 00:09:19.750
I'm curious, you know, obviously USAA is
00:09:19.760 --> 00:09:22.550
a really well- reggarded organization.
00:09:22.560 --> 00:09:25.190
Um, you spent a long time there. I think
00:09:25.200 --> 00:09:28.710
you've you've made some comments um
00:09:28.720 --> 00:09:31.350
about, you know, having worked in the
00:09:31.360 --> 00:09:32.870
industry for the time that you have.
00:09:32.880 --> 00:09:36.550
You've seen things work well and then
00:09:36.560 --> 00:09:38.949
um seeing where typical inefficiencies
00:09:38.959 --> 00:09:42.630
were. Are there any things that
00:09:42.640 --> 00:09:44.389
you know you thought were were maybe
00:09:44.399 --> 00:09:46.550
unique to USAA but now being on the
00:09:46.560 --> 00:09:47.670
other side of the table you've
00:09:47.680 --> 00:09:52.150
discovered are pretty widespread?
00:09:52.160 --> 00:09:55.350
>> Yeah, absolutely.
00:09:55.360 --> 00:09:59.110
So when I was going through at USA, the
00:09:59.120 --> 00:10:00.949
last role I had there was around
00:10:00.959 --> 00:10:02.550
modernization
00:10:02.560 --> 00:10:07.110
and I was uh to put it nicely
00:10:07.120 --> 00:10:11.269
embarrassed by some of the the
00:10:11.279 --> 00:10:13.990
processes, the slowness, the technology
00:10:14.000 --> 00:10:15.829
we had in place. We had some technology
00:10:15.839 --> 00:10:18.550
that was old and we had some processes
00:10:18.560 --> 00:10:20.949
that had been built up to work around
00:10:20.959 --> 00:10:25.590
things. some that uh for example uh my
00:10:25.600 --> 00:10:28.790
pricing teams had um some of their some
00:10:28.800 --> 00:10:31.829
of my analysts were spending more time
00:10:31.839 --> 00:10:34.870
validating data handoffs than they were
00:10:34.880 --> 00:10:37.910
doing analytical work that they'd been
00:10:37.920 --> 00:10:39.509
trained to do, right? They taken all
00:10:39.519 --> 00:10:42.550
these actuarial exams to do and yet
00:10:42.560 --> 00:10:45.030
they're checking spreadsheets. Um, and
00:10:45.040 --> 00:10:48.550
so I was embarrassed and knew we needed
00:10:48.560 --> 00:10:50.630
to fix it and and I'll tell you they're
00:10:50.640 --> 00:10:52.069
they're amazing and how they're moving
00:10:52.079 --> 00:10:54.310
forward and and getting all that fixed.
00:10:54.320 --> 00:10:56.790
But when I came to Ernex and started
00:10:56.800 --> 00:10:59.269
seeing more broadly across the industry,
00:10:59.279 --> 00:11:02.069
I saw this was not a a problem with one
00:11:02.079 --> 00:11:05.350
company. This was a problem that is
00:11:05.360 --> 00:11:07.750
across the industry, right? and and it
00:11:07.760 --> 00:11:11.430
really has built up over time and it is
00:11:11.440 --> 00:11:13.829
it it's made worse
00:11:13.839 --> 00:11:16.150
as uh I'd say for some of these more
00:11:16.160 --> 00:11:18.150
innovative companies that have had
00:11:18.160 --> 00:11:20.389
legacy technology holding them back and
00:11:20.399 --> 00:11:22.470
they have built this scaffolding of
00:11:22.480 --> 00:11:24.710
workarounds around everything they have
00:11:24.720 --> 00:11:27.590
out there that almost feels like look
00:11:27.600 --> 00:11:29.750
we're not being held back by legacy
00:11:29.760 --> 00:11:32.230
technology not counting all the tech
00:11:32.240 --> 00:11:35.269
debt and the workarounds and the side
00:11:35.279 --> 00:11:37.509
projects and the ones spreadsheet that
00:11:37.519 --> 00:11:39.990
is actually the process that is holding
00:11:40.000 --> 00:11:41.829
everything up, not the system you have
00:11:41.839 --> 00:11:44.949
in place. Those are the things that I
00:11:44.959 --> 00:11:47.030
think uh sometimes hold these companies
00:11:47.040 --> 00:11:50.230
back. Um but but yes, it is I I found
00:11:50.240 --> 00:11:52.550
that it is much it is much more an
00:11:52.560 --> 00:11:55.350
industry problem for us to solve and not
00:11:55.360 --> 00:11:57.430
just a one company or another company
00:11:57.440 --> 00:11:59.910
problem to solve.
00:11:59.920 --> 00:12:01.910
>> And then one maybe from the other side.
00:12:01.920 --> 00:12:03.509
So,
00:12:03.519 --> 00:12:06.550
did did did you have sort of a a
00:12:06.560 --> 00:12:10.069
perception of I I always hate the word
00:12:10.079 --> 00:12:12.230
vendors, but for lack of a better term,
00:12:12.240 --> 00:12:15.910
vendors uh when you were inside and now
00:12:15.920 --> 00:12:17.190
that you're on the other side of the
00:12:17.200 --> 00:12:19.030
table, you know, do you see a different
00:12:19.040 --> 00:12:21.190
reality?
00:12:21.200 --> 00:12:25.590
>> Uh yes and no. Uh so I've I I think I
00:12:25.600 --> 00:12:27.990
there's times where even within earnex,
00:12:28.000 --> 00:12:31.269
I make uh my own sales team mad at me.
00:12:31.279 --> 00:12:34.150
uh when I tell them things like no no no
00:12:34.160 --> 00:12:35.750
if I got that email I wouldn't even
00:12:35.760 --> 00:12:40.310
respond. Um and I was in the position of
00:12:40.320 --> 00:12:43.829
being an insurance executive that got
00:12:43.839 --> 00:12:46.710
tens dozens of emails every day from
00:12:46.720 --> 00:12:49.350
various vendors saying we can fix all
00:12:49.360 --> 00:12:52.550
your problems. Um, the the ones that
00:12:52.560 --> 00:12:55.190
made a difference for me back then were
00:12:55.200 --> 00:12:58.150
the ones that I, you know, knew
00:12:58.160 --> 00:12:59.910
something about me, were introduced to
00:12:59.920 --> 00:13:02.470
me by someone else that, you know, I had
00:13:02.480 --> 00:13:06.629
seen actual results from, uh, that I had
00:13:06.639 --> 00:13:08.310
any sort of a personal relationship
00:13:08.320 --> 00:13:11.829
with, or that had that one tagline that
00:13:11.839 --> 00:13:13.350
that actually caught my attention
00:13:13.360 --> 00:13:14.790
because it was a problem I was already
00:13:14.800 --> 00:13:16.150
trying to fix. not something they're
00:13:16.160 --> 00:13:17.670
trying to teach me that I had a problem
00:13:17.680 --> 00:13:19.829
with, but a problem I already had that I
00:13:19.839 --> 00:13:21.990
needed to fix. When I now that I'm on
00:13:22.000 --> 00:13:25.350
the vendor side, I see how hard uh all
00:13:25.360 --> 00:13:27.509
these sales people work to really get
00:13:27.519 --> 00:13:32.470
the word out there. Um and it is uh it
00:13:32.480 --> 00:13:35.910
is sometimes difficult um you know for
00:13:35.920 --> 00:13:39.110
various executives that are pulled into
00:13:39.120 --> 00:13:41.030
the dayto-day
00:13:41.040 --> 00:13:43.910
of just keeping the business running
00:13:43.920 --> 00:13:46.230
keeping things afloat to be able to
00:13:46.240 --> 00:13:48.949
think strategically to say not just what
00:13:48.959 --> 00:13:50.710
am I doing today but what should I be
00:13:50.720 --> 00:13:52.389
doing in the future you know if if I
00:13:52.399 --> 00:13:55.110
took away the the maze that's in front
00:13:55.120 --> 00:13:57.750
of me or this you know legacy system
00:13:57.760 --> 00:13:59.910
that's forcing ing me to do one thing or
00:13:59.920 --> 00:14:01.590
another.
00:14:01.600 --> 00:14:03.350
Does that executive have enough time to
00:14:03.360 --> 00:14:05.750
really think forward thinking about what
00:14:05.760 --> 00:14:10.069
they need to do? Uh so if I had known
00:14:10.079 --> 00:14:11.829
now a little bit more what I'd known
00:14:11.839 --> 00:14:13.189
then, I think I probably would have
00:14:13.199 --> 00:14:15.269
carved out even more time to think
00:14:15.279 --> 00:14:16.389
strategically.
00:14:16.399 --> 00:14:16.790
>> Yeah.
00:14:16.800 --> 00:14:19.670
>> Back in in those positions. Uh but but
00:14:19.680 --> 00:14:21.670
that's where you know both sides have to
00:14:21.680 --> 00:14:24.389
take a grain of salt and say you know
00:14:24.399 --> 00:14:26.150
these executives and insurance companies
00:14:26.160 --> 00:14:27.910
they've got important things they got to
00:14:27.920 --> 00:14:30.870
keep running and get done and they have
00:14:30.880 --> 00:14:33.829
to think strategically. So, you know, if
00:14:33.839 --> 00:14:36.150
we on the vendor side can make sure that
00:14:36.160 --> 00:14:38.550
they they're getting what they need when
00:14:38.560 --> 00:14:40.710
they need it and being able to to be
00:14:40.720 --> 00:14:45.189
shown not just the sales pitch, but uh
00:14:45.199 --> 00:14:47.110
what is the reality and what could what
00:14:47.120 --> 00:14:48.710
could life look like in the future if
00:14:48.720 --> 00:14:51.910
they started to make some changes.
00:14:51.920 --> 00:14:54.230
So, you know, I think one of the points
00:14:54.240 --> 00:14:57.750
you've you've made is that there's
00:14:57.760 --> 00:15:00.550
um you know, technology is one thing,
00:15:00.560 --> 00:15:02.710
but there really needs to be alignment
00:15:02.720 --> 00:15:05.670
across an organization to have effective
00:15:05.680 --> 00:15:07.189
transformation.
00:15:07.199 --> 00:15:10.470
Um, someone I've I've done some some
00:15:10.480 --> 00:15:13.269
things with over the years and insure
00:15:13.279 --> 00:15:15.189
has often used the term innovation
00:15:15.199 --> 00:15:17.350
theater. So, I guess maybe I'll I'll
00:15:17.360 --> 00:15:19.990
I'll I'll sort of copy that and say, you
00:15:20.000 --> 00:15:22.790
know, what makes for real transformation
00:15:22.800 --> 00:15:24.470
versus just sort of transformation
00:15:24.480 --> 00:15:26.710
theater.
00:15:26.720 --> 00:15:30.389
>> I I like that transformation theater. Uh
00:15:30.399 --> 00:15:34.790
many sales cycles sell the demo. It's on
00:15:34.800 --> 00:15:36.550
the implementation side that we're
00:15:36.560 --> 00:15:39.030
absorbing reality that maybe didn't make
00:15:39.040 --> 00:15:43.189
it into the slide. Right. So,
00:15:43.199 --> 00:15:45.590
you talked about ownership. I think that
00:15:45.600 --> 00:15:48.470
uh I've been through quite a few
00:15:48.480 --> 00:15:51.269
implementation projects now on both
00:15:51.279 --> 00:15:55.509
sides and you're right that if ownership
00:15:55.519 --> 00:15:58.949
is not there then implementation not
00:15:58.959 --> 00:16:00.710
only goes slow but it sometimes just
00:16:00.720 --> 00:16:03.670
doesn't even happen right you it's not
00:16:03.680 --> 00:16:05.829
at implementation that you notice it
00:16:05.839 --> 00:16:09.590
it's it's 3 to 6 months into a project
00:16:09.600 --> 00:16:11.189
where
00:16:11.199 --> 00:16:13.350
somebody has to make a decision to kill
00:16:13.360 --> 00:16:15.990
a customization
00:16:16.000 --> 00:16:17.350
something that you know the company's
00:16:17.360 --> 00:16:20.310
built up over time or force a business
00:16:20.320 --> 00:16:23.829
unit to change how it works. And if
00:16:23.839 --> 00:16:26.069
there's no single owner with the real
00:16:26.079 --> 00:16:28.470
authority to actually get that done,
00:16:28.480 --> 00:16:31.749
then it's not going to happen. Most
00:16:31.759 --> 00:16:33.269
programs are not going to fail at the go
00:16:33.279 --> 00:16:34.870
live, right? They're going to fail three
00:16:34.880 --> 00:16:36.790
months before it. They're going to when
00:16:36.800 --> 00:16:39.590
somebody has to make that unpopular call
00:16:39.600 --> 00:16:41.590
and there's nobody with authority to
00:16:41.600 --> 00:16:43.990
actually make it happen. those shadow
00:16:44.000 --> 00:16:46.470
legacy processes are just going to keep
00:16:46.480 --> 00:16:50.069
going. Um, nobody's going to kill their
00:16:50.079 --> 00:16:52.629
own uh customization that they put in
00:16:52.639 --> 00:16:54.550
that they own that they they're very
00:16:54.560 --> 00:16:56.150
proud of, right? They need somebody from
00:16:56.160 --> 00:17:00.389
the top with the vision to say hold on
00:17:00.399 --> 00:17:02.470
uh we have to think about this long
00:17:02.480 --> 00:17:05.429
term, not just in small buckets or what
00:17:05.439 --> 00:17:08.150
individuals do throughout the process.
00:17:08.160 --> 00:17:11.669
The the other place where I've seen uh
00:17:11.679 --> 00:17:15.510
leadership really come into this has to
00:17:15.520 --> 00:17:20.069
do with uh parallel processes. Uh
00:17:20.079 --> 00:17:23.270
anytime you're going to put uh you know
00:17:23.280 --> 00:17:27.510
modernization into effect. Uh it's not
00:17:27.520 --> 00:17:29.110
going to be where you build it all on
00:17:29.120 --> 00:17:32.230
the side and then people come into work
00:17:32.240 --> 00:17:33.430
one day and everything will be
00:17:33.440 --> 00:17:36.549
different. Uh you're building it up.
00:17:36.559 --> 00:17:38.950
you're running parallel processes. Uh
00:17:38.960 --> 00:17:40.789
you're testing and learning into it. You
00:17:40.799 --> 00:17:42.070
maybe, you know, are doing it for one
00:17:42.080 --> 00:17:44.549
line of business or a few states within
00:17:44.559 --> 00:17:47.190
a line of business or for one team and
00:17:47.200 --> 00:17:51.270
not other teams. Um what that causes is
00:17:51.280 --> 00:17:54.710
it it's a risk to the project because
00:17:54.720 --> 00:17:56.789
you're causing real fatigue in the
00:17:56.799 --> 00:17:59.029
people that are actually saying, "You
00:17:59.039 --> 00:18:00.789
told me my job was going to get better
00:18:00.799 --> 00:18:03.669
as it was going to get easier.
00:18:03.679 --> 00:18:05.350
Yet this is not easier and this is not
00:18:05.360 --> 00:18:07.430
better. I'm running two processes at
00:18:07.440 --> 00:18:08.390
this point.
00:18:08.400 --> 00:18:08.710
>> Yeah.
00:18:08.720 --> 00:18:11.750
>> And there's burnout. So without
00:18:11.760 --> 00:18:13.350
leadership there to set the right
00:18:13.360 --> 00:18:16.230
expectations and make sure that things
00:18:16.240 --> 00:18:18.230
are progressing and to watch for that
00:18:18.240 --> 00:18:22.870
risk, then it can it can kill a project
00:18:22.880 --> 00:18:24.789
or at least kill momentum within a
00:18:24.799 --> 00:18:28.310
project.
00:18:28.320 --> 00:18:31.029
I I'm also curious given your background
00:18:31.039 --> 00:18:34.150
as an actuary, does that influence
00:18:34.160 --> 00:18:36.870
your approach in in in sort of how you
00:18:36.880 --> 00:18:39.909
advise clients and then also how you run
00:18:39.919 --> 00:18:42.710
projects?
00:18:42.720 --> 00:18:45.830
>> Oh, as an actuary, uh so I I'll tell you
00:18:45.840 --> 00:18:47.830
as as an actuary, what it does is it
00:18:47.840 --> 00:18:49.750
makes me love spreadsheets and numbers
00:18:49.760 --> 00:18:53.029
and data. Um which, uh I I say tongue
00:18:53.039 --> 00:18:55.909
and cheek, but it's true. Um I I am
00:18:55.919 --> 00:18:59.510
responsible uh you know for many many
00:18:59.520 --> 00:19:01.750
I'll say userdefined applications to
00:19:01.760 --> 00:19:04.630
work around systems in the past because
00:19:04.640 --> 00:19:06.549
I want to get things done and I've built
00:19:06.559 --> 00:19:09.669
them on my own on the side. So it it
00:19:09.679 --> 00:19:12.310
helps me I'll say within projects to
00:19:12.320 --> 00:19:15.590
sympathize with these teams that say but
00:19:15.600 --> 00:19:17.750
look at this amazing thing we've done.
00:19:17.760 --> 00:19:20.470
Uh but but as a leader, we have to be
00:19:20.480 --> 00:19:23.430
able to put some of that aside um and
00:19:23.440 --> 00:19:26.470
and be able to say where do we start?
00:19:26.480 --> 00:19:28.789
Where do we end? Where are our success
00:19:28.799 --> 00:19:31.430
metrics? Let's put some actual
00:19:31.440 --> 00:19:33.110
governance in place. Let's put some
00:19:33.120 --> 00:19:36.150
schedules, process checks to get things
00:19:36.160 --> 00:19:39.750
done. Uh so, you know, yes, being an
00:19:39.760 --> 00:19:41.510
actuary makes me love spreadsheets. It
00:19:41.520 --> 00:19:45.270
makes me love data. uh but uh growing
00:19:45.280 --> 00:19:47.909
into a leader it's helped me see how we
00:19:47.919 --> 00:19:49.430
have to have that bigger picture in
00:19:49.440 --> 00:19:50.549
mind.
00:19:50.559 --> 00:19:52.789
>> Yeah. Just one one quick question on
00:19:52.799 --> 00:19:56.950
that. So it sounds like um you know that
00:19:56.960 --> 00:19:59.909
with with all the AI tools available now
00:19:59.919 --> 00:20:02.150
that that you know the term builder is a
00:20:02.160 --> 00:20:04.310
big one. Uh you know individuals can be
00:20:04.320 --> 00:20:07.270
builders now. So, it sounds like,
00:20:07.280 --> 00:20:08.710
you know, whether it's spreadsheets or
00:20:08.720 --> 00:20:10.710
or other
00:20:10.720 --> 00:20:12.630
um sort of personal tools you developed,
00:20:12.640 --> 00:20:14.390
it sounds like you've been kind of that
00:20:14.400 --> 00:20:17.830
that that builder profile for a while
00:20:17.840 --> 00:20:21.669
now. Has the availability of of of sort
00:20:21.679 --> 00:20:23.350
of tools to use personally, not
00:20:23.360 --> 00:20:25.350
necessarily, you know, organizationwide
00:20:25.360 --> 00:20:27.350
big AI initiatives, but tools available
00:20:27.360 --> 00:20:29.590
to use personally change the way you you
00:20:29.600 --> 00:20:32.630
work or do any of that building at all?
00:20:32.640 --> 00:20:36.149
>> It does. Um, so
00:20:36.159 --> 00:20:37.750
you have to be careful when you're doing
00:20:37.760 --> 00:20:39.190
anything for work, right? Because
00:20:39.200 --> 00:20:40.470
especially in the the governed
00:20:40.480 --> 00:20:42.070
environment we live, it has to make sure
00:20:42.080 --> 00:20:43.909
that the data is protected and all the
00:20:43.919 --> 00:20:47.029
rest of that. So I will say some at work
00:20:47.039 --> 00:20:49.750
and some in my personal life, I love the
00:20:49.760 --> 00:20:52.870
availability of vibe coding, right?
00:20:52.880 --> 00:20:56.630
uh it if I have to get things done now
00:20:56.640 --> 00:20:59.510
and I want something pretty uh you know
00:20:59.520 --> 00:21:02.230
a nice way to do it with vibe coding I
00:21:02.240 --> 00:21:03.990
can usually put something together
00:21:04.000 --> 00:21:07.190
pretty darn fast. Uh if I want to use
00:21:07.200 --> 00:21:09.909
that though in an insurance environment
00:21:09.919 --> 00:21:11.350
uh that doesn't that's where it doesn't
00:21:11.360 --> 00:21:13.430
fly right if I'm building a prototype to
00:21:13.440 --> 00:21:15.909
say art of the possible that's one thing
00:21:15.919 --> 00:21:18.789
but if I want long-term governance if I
00:21:18.799 --> 00:21:21.350
want uh you know assurityity that it'll
00:21:21.360 --> 00:21:23.990
work not just for my main use case but
00:21:24.000 --> 00:21:28.549
extensible to to others uh
00:21:28.559 --> 00:21:30.549
I am happy to build a prototype with
00:21:30.559 --> 00:21:34.710
vibe coding um and show it to my IT team
00:21:34.720 --> 00:21:37.029
show it to my developers and say, "No,
00:21:37.039 --> 00:21:39.190
no, no. I'm not just going to write up a
00:21:39.200 --> 00:21:40.870
description about what I needed to do. I
00:21:40.880 --> 00:21:42.950
can show you here's what it does in the
00:21:42.960 --> 00:21:45.830
best use case." But I understand that on
00:21:45.840 --> 00:21:48.870
my side, um, that there's too much at
00:21:48.880 --> 00:21:50.870
stake within insurance companies to
00:21:50.880 --> 00:21:53.830
trust somebody like me to vibe code
00:21:53.840 --> 00:21:56.070
something up to use on a long-term
00:21:56.080 --> 00:21:57.190
basis.
00:21:57.200 --> 00:22:00.310
>> Yeah. So, I want to make sure we we get
00:22:00.320 --> 00:22:02.149
to
00:22:02.159 --> 00:22:06.149
um the agents that that uh Ernex
00:22:06.159 --> 00:22:09.430
launched uh earlier this year, I think a
00:22:09.440 --> 00:22:13.590
month or or so ago. Um could you share a
00:22:13.600 --> 00:22:15.590
little bit more about you know what you
00:22:15.600 --> 00:22:18.230
guys are are offering and you know what
00:22:18.240 --> 00:22:21.190
what you see agents actually doing in a
00:22:21.200 --> 00:22:23.510
carrier today and you know is it
00:22:23.520 --> 00:22:25.110
complementing what people are doing? Is
00:22:25.120 --> 00:22:27.830
it yeah, you know, replacing what people
00:22:27.840 --> 00:22:30.070
are doing and and how it's kind of
00:22:30.080 --> 00:22:32.580
fitting into the to the workflows.
00:22:32.590 --> 00:22:33.029
[snorts]
00:22:33.039 --> 00:22:36.549
>> Yeah, that's a great question. So, so
00:22:36.559 --> 00:22:38.950
let me start by just saying in general
00:22:38.960 --> 00:22:43.029
um our view of how AI should be used
00:22:43.039 --> 00:22:46.549
within insurance is to enable expertise,
00:22:46.559 --> 00:22:49.430
right? uh there's probably going to be
00:22:49.440 --> 00:22:53.270
some jobs that uh you know down the line
00:22:53.280 --> 00:22:55.750
could be are are going to be impacted,
00:22:55.760 --> 00:22:58.310
you know, but the reality is we're not
00:22:58.320 --> 00:22:59.830
seeing companies and we probably
00:22:59.840 --> 00:23:01.270
shouldn't be seeing companies that are
00:23:01.280 --> 00:23:02.870
saying I'm going to replace 20% of my
00:23:02.880 --> 00:23:06.710
workforce with AI. That's that's not the
00:23:06.720 --> 00:23:09.029
right stage we're at right now. where we
00:23:09.039 --> 00:23:11.830
are now is we need to be saying for the
00:23:11.840 --> 00:23:13.350
experts that are out there, how do we
00:23:13.360 --> 00:23:15.110
take some of the lowhanging fruit, the
00:23:15.120 --> 00:23:16.390
things that they're doing within their
00:23:16.400 --> 00:23:19.750
job and make it easier. Uh companies
00:23:19.760 --> 00:23:22.149
have innovation all the time that sits
00:23:22.159 --> 00:23:24.149
on the shelf because they just can't get
00:23:24.159 --> 00:23:27.990
it into production. AI is no different.
00:23:28.000 --> 00:23:32.230
So what we at Ernex have really uh are
00:23:32.240 --> 00:23:33.750
launching right now and have already
00:23:33.760 --> 00:23:35.270
launched earlier this year and starting
00:23:35.280 --> 00:23:38.230
to use some of our customers is the
00:23:38.240 --> 00:23:40.710
concept of uh AIOS. And while I won't
00:23:40.720 --> 00:23:43.830
get into the the full uh talk track
00:23:43.840 --> 00:23:46.870
here, what I will tell you is that uh we
00:23:46.880 --> 00:23:50.789
allow for verticalized contextualized AI
00:23:50.799 --> 00:23:53.270
within the earnex and uh insurance
00:23:53.280 --> 00:23:55.990
framework to be embedded in the workflow
00:23:56.000 --> 00:23:58.549
of what it is that our customers are
00:23:58.559 --> 00:24:01.750
doing today. helping them actually see
00:24:01.760 --> 00:24:04.390
um how different parts of the flow and
00:24:04.400 --> 00:24:07.590
the process actually can be governed so
00:24:07.600 --> 00:24:09.750
that when as that regulation is evolving
00:24:09.760 --> 00:24:12.070
that it's all controlled. You know
00:24:12.080 --> 00:24:13.750
exactly what's there, how it's being
00:24:13.760 --> 00:24:16.230
done, how it's being operationalized
00:24:16.240 --> 00:24:17.990
um and it's actually helping the people
00:24:18.000 --> 00:24:19.269
at the point where they're making
00:24:19.279 --> 00:24:21.750
decisions. I gave an example a little
00:24:21.760 --> 00:24:23.269
earlier as we're in our discussion
00:24:23.279 --> 00:24:24.870
around those call center
00:24:24.880 --> 00:24:27.909
representatives. Um, we had one of our
00:24:27.919 --> 00:24:31.669
customers tell us that the the tenure
00:24:31.679 --> 00:24:34.230
the average tenure for call center
00:24:34.240 --> 00:24:35.990
representatives is about 12 months at
00:24:36.000 --> 00:24:38.149
their company and it takes 18 months to
00:24:38.159 --> 00:24:40.549
get them fully trained. Well, that
00:24:40.559 --> 00:24:43.190
doesn't work. And so, if you can instead
00:24:43.200 --> 00:24:46.470
be able to embed this verticalized AI
00:24:46.480 --> 00:24:49.830
within to do something like explain uh
00:24:49.840 --> 00:24:52.470
why premium changed, right? that that
00:24:52.480 --> 00:24:55.269
makes it not only so that the training
00:24:55.279 --> 00:24:57.990
burden that's on this is shrunk so that
00:24:58.000 --> 00:25:01.350
you're getting them up and going faster,
00:25:01.360 --> 00:25:03.590
but I also wonder while I haven't seen
00:25:03.600 --> 00:25:06.310
numbers on this if it could extend that
00:25:06.320 --> 00:25:08.310
12 months to something longer by making
00:25:08.320 --> 00:25:11.269
somebody's job that much easier and that
00:25:11.279 --> 00:25:13.269
much better. Right? Not only is the
00:25:13.279 --> 00:25:15.990
company having good consistent
00:25:16.000 --> 00:25:18.070
explanations from customer to customer
00:25:18.080 --> 00:25:21.110
that match reality, uh, but for a call
00:25:21.120 --> 00:25:23.909
center agent who is juggling everything
00:25:23.919 --> 00:25:25.990
under the sun to be able to have
00:25:26.000 --> 00:25:28.070
something brought directly to them that
00:25:28.080 --> 00:25:30.470
it has access to all the operational
00:25:30.480 --> 00:25:34.070
decisions, the customer's profile, the
00:25:34.080 --> 00:25:36.070
right constraints put around it around
00:25:36.080 --> 00:25:38.310
what language should be there um, and
00:25:38.320 --> 00:25:40.149
served up to them right when they need
00:25:40.159 --> 00:25:42.710
it in the tools they're already
00:25:42.720 --> 00:25:45.510
That's gold and that's what makes uh
00:25:45.520 --> 00:25:47.350
makes things actually work and better
00:25:47.360 --> 00:25:49.590
and that's what'll flow through then to
00:25:49.600 --> 00:25:52.470
your CFO's bottom line.
00:25:52.480 --> 00:25:54.950
>> So I don't even know the right time
00:25:54.960 --> 00:25:57.269
frame to use anymore. You know in the
00:25:57.279 --> 00:25:58.789
past I would have said three years out
00:25:58.799 --> 00:26:00.950
but things are moving so quickly. I I
00:26:00.960 --> 00:26:03.590
don't know what what what to use. But
00:26:03.600 --> 00:26:09.669
you know what what is a team actually
00:26:09.679 --> 00:26:10.870
you know and maybe you can talk about
00:26:10.880 --> 00:26:13.430
individual personas but what is a team
00:26:13.440 --> 00:26:16.310
actually doing sort of all day once this
00:26:16.320 --> 00:26:18.870
is really working and and embedded in
00:26:18.880 --> 00:26:21.669
our organizations.
00:26:21.679 --> 00:26:24.950
Yeah, I I don't think that the
00:26:24.960 --> 00:26:28.470
the steps needed to price an insurance
00:26:28.480 --> 00:26:32.149
policy or to determine the right, you
00:26:32.159 --> 00:26:34.070
know, underwriting analytics and, you
00:26:34.080 --> 00:26:36.070
know, knock criteria, underwriting
00:26:36.080 --> 00:26:37.430
guidance. I don't think that the
00:26:37.440 --> 00:26:39.750
fundamentals around that changes. I
00:26:39.760 --> 00:26:42.630
think what does change is that the
00:26:42.640 --> 00:26:45.909
decisions that are being made, the speed
00:26:45.919 --> 00:26:48.870
it takes to get things done. uh for
00:26:48.880 --> 00:26:51.269
example you know a data scientist that's
00:26:51.279 --> 00:26:55.269
building a loss cost model uh for you
00:26:55.279 --> 00:26:57.750
know insurance pricing
00:26:57.760 --> 00:27:00.630
when I did some of the first ones back
00:27:00.640 --> 00:27:03.190
in the early 2000s it took me forever it
00:27:03.200 --> 00:27:05.830
took me months to get it done because I
00:27:05.840 --> 00:27:07.269
didn't know what I was doing the
00:27:07.279 --> 00:27:09.669
technology was new
00:27:09.679 --> 00:27:13.110
nowadays there's actually and I'll tell
00:27:13.120 --> 00:27:14.950
you earnex has some of this embedded in
00:27:14.960 --> 00:27:17.750
our data science platform abilities for
00:27:17.760 --> 00:27:19.750
AI to jump in and say, "Let me tell you
00:27:19.760 --> 00:27:21.590
the difference between two versions of a
00:27:21.600 --> 00:27:24.710
model. Let me suggest variables that you
00:27:24.720 --> 00:27:27.430
should consider. Let me look at the fits
00:27:27.440 --> 00:27:29.830
that you have within there or the the
00:27:29.840 --> 00:27:31.909
bucketing of your different variables,
00:27:31.919 --> 00:27:33.990
smoothing capabilities." I mean, all of
00:27:34.000 --> 00:27:35.830
these different things that a data
00:27:35.840 --> 00:27:39.350
scientist has to consider. Now, we're
00:27:39.360 --> 00:27:40.950
not taking away the need for their
00:27:40.960 --> 00:27:43.029
expertise, but what we are doing is
00:27:43.039 --> 00:27:44.710
making it so that as they're doing their
00:27:44.720 --> 00:27:46.950
job, it's smoother along the way. And I
00:27:46.960 --> 00:27:48.310
think that's across the board with all
00:27:48.320 --> 00:27:50.070
these different experts that we're
00:27:50.080 --> 00:27:53.029
looking at that, you know, the value
00:27:53.039 --> 00:27:55.750
added work that they do today will be
00:27:55.760 --> 00:27:58.389
there and will be magnified. It's just
00:27:58.399 --> 00:28:00.230
the stuff that takes away from them
00:28:00.240 --> 00:28:02.230
actually using some of that knowledge
00:28:02.240 --> 00:28:04.389
and ability they have that needs to fall
00:28:04.399 --> 00:28:07.350
by the wayside. uh taking a spreadsheet
00:28:07.360 --> 00:28:08.950
and double-checking it between, you
00:28:08.960 --> 00:28:11.110
know, two different screens or copying
00:28:11.120 --> 00:28:15.190
and pasting data from an email into an
00:28:15.200 --> 00:28:17.909
operational system. That that shouldn't
00:28:17.919 --> 00:28:19.590
have to be somebody's job. It's not
00:28:19.600 --> 00:28:22.149
something somebody enjoys doing, and it
00:28:22.159 --> 00:28:23.350
should be something that we're able to
00:28:23.360 --> 00:28:26.230
hand to an AI and make it help uh where
00:28:26.240 --> 00:28:28.470
it's supposed to.
00:28:28.480 --> 00:28:30.310
>> So, Aaron, I really appreciate your
00:28:30.320 --> 00:28:32.230
insights. Is there anything we haven't
00:28:32.240 --> 00:28:33.750
touched on today that you'd like to
00:28:33.760 --> 00:28:37.909
cover before we wrap up?
00:28:37.919 --> 00:28:41.830
>> Maybe just at the end I I'll say that uh
00:28:41.840 --> 00:28:44.630
you know I through my career I have seen
00:28:44.640 --> 00:28:49.510
where uh legacy systems have really
00:28:49.520 --> 00:28:52.070
slowed companies down and while yes
00:28:52.080 --> 00:28:53.830
there are workarounds those are still
00:28:53.840 --> 00:28:57.029
another form of slowing us down. I I
00:28:57.039 --> 00:28:59.029
think as an industry we really need to
00:28:59.039 --> 00:29:01.510
take a hard look at what the art of the
00:29:01.520 --> 00:29:03.590
possible is in the future and we're
00:29:03.600 --> 00:29:05.830
getting pressure from both sides from
00:29:05.840 --> 00:29:07.750
you know legacy systems slowing us down
00:29:07.760 --> 00:29:09.669
from new technology with AI that's out
00:29:09.679 --> 00:29:12.070
there that can allow us to speed up and
00:29:12.080 --> 00:29:14.070
where we meet in the middle is going to
00:29:14.080 --> 00:29:16.149
be how we take our industry forward and
00:29:16.159 --> 00:29:17.510
I think it's important for us to give
00:29:17.520 --> 00:29:20.549
that the right amount of thought and uh
00:29:20.559 --> 00:29:22.470
time and effort to make sure that that
00:29:22.480 --> 00:29:24.389
happens and that we as an industry can
00:29:24.399 --> 00:29:27.510
actually move forward in the right
00:29:27.520 --> 00:29:31.029
Yeah, I mean I uh I do think sometimes
00:29:31.039 --> 00:29:33.029
there's there's too much do and not
00:29:33.039 --> 00:29:35.350
enough think. um you know sort of ready
00:29:35.360 --> 00:29:38.389
fire aim ready fire aim situation
00:29:38.399 --> 00:29:42.070
because you know and I do think to your
00:29:42.080 --> 00:29:45.990
point before you know insurance it you
00:29:46.000 --> 00:29:47.430
know it's an industry that's all about
00:29:47.440 --> 00:29:49.510
risk management so it tends to be
00:29:49.520 --> 00:29:51.029
conservative
00:29:51.039 --> 00:29:53.909
but you know inevitably it does face a
00:29:53.919 --> 00:29:55.510
lot of the same pressures that that
00:29:55.520 --> 00:29:58.070
every other industry does um you know do
00:29:58.080 --> 00:30:01.190
more do faster do more efficiently do
00:30:01.200 --> 00:30:07.669
cheaper um And um and uh
00:30:07.679 --> 00:30:12.549
you know it can and it can create uh uh
00:30:12.559 --> 00:30:14.710
you know in this this environment uh
00:30:14.720 --> 00:30:17.190
sort of that that that FOMO and and and
00:30:17.200 --> 00:30:20.470
and people can send them off down a path
00:30:20.480 --> 00:30:22.789
that that doesn't always make the most
00:30:22.799 --> 00:30:24.230
sense just because they feel like they
00:30:24.240 --> 00:30:26.149
have to do something.
00:30:26.159 --> 00:30:27.990
>> Yeah. Very true.
00:30:28.000 --> 00:30:28.389
>> Yeah.
00:30:28.399 --> 00:30:30.950
>> Very true. Well, Aaron Wright, VP of
00:30:30.960 --> 00:30:33.990
Professional Services Americas at Ernex,
00:30:34.000 --> 00:30:36.070
thanks so much for joining me on this
00:30:36.080 --> 00:30:38.389
episode of the Insure Tech Leadership
00:30:38.399 --> 00:30:39.750
Podcast.
00:30:39.760 --> 00:30:43.120
>> Thank you, Josh. I love
