1
00:00:01,357 --> 00:00:05,897
Welcome to Reinventing Professionals,
a podcast hosted by industry analyst
2
00:00:05,977 --> 00:00:10,777
Ari Kaplan, which shares ideas,
guidance, and perspectives from market
3
00:00:10,837 --> 00:00:15,118
leaders shaping the next generation
of legal and professional services.
4
00:00:19,068 --> 00:00:24,628
This is Ari Kaplan, and I'm speaking
today with Frédéric Beylier, the CEO
5
00:00:24,708 --> 00:00:29,108
of Questel, a provider of intellectual
property solutions, including AI-powered
6
00:00:29,669 --> 00:00:32,608
modular software and expert services.
7
00:00:33,088 --> 00:00:33,448
Hi, Fred.
8
00:00:33,488 --> 00:00:33,869
How are you?
9
00:00:34,648 --> 00:00:35,248
Hi, Ari.
10
00:00:35,328 --> 00:00:36,228
Thanks for having me.
11
00:00:36,408 --> 00:00:37,528
Great to speak to you today.
12
00:00:38,108 --> 00:00:39,428
It's my privilege to speak with you.
13
00:00:39,469 --> 00:00:43,228
So tell us about your background
and your role at Questel.
14
00:00:44,548 --> 00:00:47,529
I've been running companies in the
tech space for the past 20 years now.
15
00:00:48,388 --> 00:00:51,348
Initially in the security
technology space.
16
00:00:51,439 --> 00:00:56,289
Then I moved also to the mobility tech
mobility space, the urban mobility.
17
00:00:57,149 --> 00:00:58,970
And now I'm in IP and I love it.
18
00:00:58,990 --> 00:01:00,890
There are some similarities, of course.
19
00:01:01,030 --> 00:01:04,449
Most of the time I'm getting in industry
about, there's a kind of transformation
20
00:01:04,490 --> 00:01:10,609
ongoing, so at the time of the security
technology was digital, and IoT, and
21
00:01:10,709 --> 00:01:15,430
now it's AI for IP, so really exciting
and trying to build the industry leaders
22
00:01:15,451 --> 00:01:19,121
in, in their space, of course, that's
the intent is really to try to build
23
00:01:19,221 --> 00:01:20,601
industry leaders in the long run.
24
00:01:20,691 --> 00:01:23,691
How is AI changing the IP sector?
25
00:01:25,378 --> 00:01:26,308
It has changed already.
26
00:01:26,458 --> 00:01:31,098
We did a report on AI adoption
and IP, and fifty percent of the
27
00:01:31,138 --> 00:01:34,878
respondents are declaring that
they're using AI at least once a day.
28
00:01:35,228 --> 00:01:38,028
Of course it's AI assistance,
so it's just the start of it.
29
00:01:39,128 --> 00:01:43,018
But we believe that in enterprise
as a whole and not IP, but
30
00:01:43,038 --> 00:01:46,468
enterprise the rate is more around
twenty to twenty-five percent.
31
00:01:46,548 --> 00:01:50,388
So having respondents in the IP
space answering fifty-plus percent
32
00:01:50,968 --> 00:01:53,548
at least once a day, it's not bad.
33
00:01:54,108 --> 00:01:55,168
It's pretty advanced.
34
00:01:55,558 --> 00:01:59,578
It's about AI assistance, the way you
work, the way you get a summary, the
35
00:01:59,638 --> 00:02:02,248
way you get an an opinion or you prompt.
36
00:02:02,928 --> 00:02:06,858
It can summarize a lot of data
information, and some of the tools are
37
00:02:06,878 --> 00:02:11,178
already providing this, Which are really
AI native tools like our tool in Qthena.
38
00:02:11,238 --> 00:02:14,128
Tomorrow we're gonna see more and
more agents coming up and that's what
39
00:02:14,168 --> 00:02:18,628
it's about now and helping people
executing tasks, and multiplying and
40
00:02:18,658 --> 00:02:21,848
leveraging the power of agents to
do the work on behalf of the people.
41
00:02:21,938 --> 00:02:26,938
What are the emerging challenges
that brand owners and innovators
42
00:02:26,978 --> 00:02:28,758
are facing in the current climate?
43
00:02:30,328 --> 00:02:35,488
Multiple because, of course, when we come
to innovations it's always more complex.
44
00:02:35,548 --> 00:02:36,608
You need to handle AI.
45
00:02:36,688 --> 00:02:41,258
So there are native AI players that
are used to this, but you also have
46
00:02:41,358 --> 00:02:46,538
a complexity in the way you to some
extent patent your ideas as well.
47
00:02:46,588 --> 00:02:51,218
Now it's AI open source technology and
you need to be really strong on the way
48
00:02:51,378 --> 00:02:53,358
you patent your technologies as well.
49
00:02:53,408 --> 00:02:58,648
Today there are key challenges which
are, having a homogenous adoption
50
00:02:58,688 --> 00:03:00,287
of AI, around the innovators.
51
00:03:00,388 --> 00:03:03,318
And at the same time it's an adoption
and you want a change to happen,
52
00:03:03,398 --> 00:03:06,289
but you have key constraints,
do you have proprietary data?
53
00:03:06,549 --> 00:03:07,289
Who's the data?
54
00:03:08,089 --> 00:03:09,309
Who's belonging the data?
55
00:03:09,849 --> 00:03:12,109
Do you have security at
the right level as well?
56
00:03:13,049 --> 00:03:16,009
Do you have the right ecosystem
around you in terms of integration?
57
00:03:16,368 --> 00:03:21,988
So there are a few technical complexities
when it comes to challenges of owners and
58
00:03:22,069 --> 00:03:26,368
innovators, brand owners and innovators
when it comes to AI, which are pretty big
59
00:03:26,428 --> 00:03:31,549
because you can jump into AI, but you're
also in the open world, I would say.
60
00:03:31,658 --> 00:03:35,438
And using classical LLM, you put
part of your innovation, part of your
61
00:03:36,459 --> 00:03:40,668
secret sauce, in the open world, and
that's also some critical aspects.
62
00:03:40,719 --> 00:03:43,559
And the way you scale AI is not
gonna be the same if your structure
63
00:03:43,600 --> 00:03:45,979
in term of data before that or not.
64
00:03:45,999 --> 00:03:48,100
It's really different the
way you do it as well.
65
00:03:48,179 --> 00:03:49,899
It's the how that is important.
66
00:03:49,989 --> 00:03:55,369
How is Questel addressing the risks
and opportunities associated with AI?
67
00:03:56,009 --> 00:03:59,529
We see 200% opportunities more
than risk to be honest with you.
68
00:04:00,110 --> 00:04:03,989
We see it as an opportunity because
it's an accelerator for innovation.
69
00:04:04,979 --> 00:04:06,850
We have a few AI-native products.
70
00:04:07,629 --> 00:04:12,189
Some of our more software-based product
are also now embedded with AI as well.
71
00:04:12,210 --> 00:04:12,771
We're using it.
72
00:04:12,811 --> 00:04:18,199
We have developed our own AI LLM that
we embed into our products as well.
73
00:04:18,699 --> 00:04:19,799
That's on the product side.
74
00:04:19,859 --> 00:04:21,099
And we're moving to agentic.
75
00:04:21,150 --> 00:04:25,531
We're now launching some agentic solution,
especially in our Qthena, which is really
76
00:04:25,590 --> 00:04:29,950
a kind of AI-native cockpit helping,
patent attorneys or trademark attorneys
77
00:04:29,970 --> 00:04:36,261
to draft patents, summarize patents- do
checks get all the public information.
78
00:04:36,361 --> 00:04:39,701
They can even add their proprietary
information inside it, so it's really
79
00:04:39,801 --> 00:04:43,501
a kind of cockpit for the patent
attorney to gain approximately 40%
80
00:04:43,521 --> 00:04:44,701
of their time, so that's massive.
81
00:04:45,441 --> 00:04:48,571
For us it's really an opportunity from
a business development standpoint.
82
00:04:49,271 --> 00:04:53,451
It's an opportunity as well from
an efficiency, internal efficiency,
83
00:04:53,531 --> 00:04:58,171
because the more we adopt AI internally,
the more we map our processes and
84
00:04:58,251 --> 00:05:01,591
we use AI, the more we automate our
processes, the more we are efficient.
85
00:05:02,471 --> 00:05:05,141
So we can spare time, on
more value-added task.
86
00:05:05,251 --> 00:05:06,411
Now, it's also a risk.
87
00:05:07,161 --> 00:05:10,530
The risk is more on the big LLMs
because everybody thinks they
88
00:05:10,621 --> 00:05:13,341
can do everything with the big
LLM, which is not exactly true.
89
00:05:13,761 --> 00:05:18,631
The reality of it is depending on
the depth of the work you need to do,
90
00:05:18,791 --> 00:05:21,331
you want to do the data is critical.
91
00:05:21,431 --> 00:05:25,311
So you want to have access at the same
data in an LLM than in other areas, and
92
00:05:25,411 --> 00:05:30,741
es-especially when you're deep into a
vertical like IP the data you can provide
93
00:05:30,791 --> 00:05:35,461
is much more precise and accurate and
as well qualified than your current LLM.
94
00:05:36,381 --> 00:05:38,301
And also, LLMs have
hallucinations as well.
95
00:05:38,601 --> 00:05:39,721
So you need to control that.
96
00:05:39,781 --> 00:05:42,441
You need to be on top
of the results you get.
97
00:05:42,541 --> 00:05:44,541
We see LLMs a-as a competitor.
98
00:05:44,621 --> 00:05:48,821
It's a key enabler for us as well
because we're using them and we combine
99
00:05:48,831 --> 00:05:51,081
them with our own AI we've developed.
100
00:05:51,941 --> 00:05:56,071
But also it's a competition for us
because we look at what they do and how
101
00:05:56,101 --> 00:06:03,237
they disrupt potentially the usage. And
they have for them huge fundraising,
102
00:06:03,317 --> 00:06:05,477
they have for them huge marketing spend.
103
00:06:06,357 --> 00:06:10,537
Now what we have for us is
proprietary data in IP, so
104
00:06:10,697 --> 00:06:12,077
really qualified and enriched.
105
00:06:12,937 --> 00:06:16,317
We have for us a deep
vertical functional knowledge.
106
00:06:17,177 --> 00:06:20,816
We are also, executing some
critical tasks for our customers.
107
00:06:21,197 --> 00:06:23,677
We are paying annuities
for patent annuities.
108
00:06:23,737 --> 00:06:25,917
We are renewing marks on
behalf of our customers.
109
00:06:26,507 --> 00:06:30,647
We're doing a lot of depth of
services, but also the functional
110
00:06:31,447 --> 00:06:35,447
know-how is way deeper than, of
course, what you can find in LLM.
111
00:06:35,507 --> 00:06:38,807
So that's what we have for us,
and we need to go deeper and
112
00:06:38,847 --> 00:06:42,987
deeper into our expertise, of
course, to, to be differentiated
113
00:06:43,047 --> 00:06:44,566
vis-à-vis the LLM for our customers.
114
00:06:45,987 --> 00:06:48,987
You mentioned that Questel recently
released benchmarking research
115
00:06:49,027 --> 00:06:52,147
on the adoption- Yeah … of
AI tools across the IP sector.
116
00:06:52,307 --> 00:06:59,687
What did you identify as the main
use cases of AI for IP, and where
117
00:07:00,007 --> 00:07:03,807
are IP professionals realizing
the greatest benefits from AI?
118
00:07:05,087 --> 00:07:08,307
It's really about AI assistance
and how do I gain time today.
119
00:07:09,287 --> 00:07:12,147
And that's most of the
outcome we see today.
120
00:07:12,187 --> 00:07:15,267
Everybody now acknowledges that
AI is gonna change the way I work.
121
00:07:15,327 --> 00:07:18,307
It's seventy-four percent now
acknowledging this across the
122
00:07:18,327 --> 00:07:20,087
IP space, so that's significant.
123
00:07:21,097 --> 00:07:25,297
It's a matter of efficiency being
augmented, I would say, like more
124
00:07:25,317 --> 00:07:26,657
productive and more efficient.
125
00:07:27,737 --> 00:07:31,297
Tomorrow everybody is not seeing
the agentic piece because not
126
00:07:31,357 --> 00:07:35,197
everybody's advanced as as some of
the players, and especially us, we
127
00:07:35,217 --> 00:07:38,737
are tech providers, so we see the
agentic, we develop it, we launch it.
128
00:07:38,827 --> 00:07:42,397
But it's gonna also change the value
chain in IP, I would say, because
129
00:07:42,477 --> 00:07:46,817
some basic services that you used
to call a third-party provider, you
130
00:07:46,877 --> 00:07:49,867
can do it through agents now, so
it's really different in the way
131
00:07:49,887 --> 00:07:50,947
it's gonna change the value chain.
132
00:07:51,037 --> 00:07:55,597
For anybody in the IP space, lawyers,
they need to bring more added value to
133
00:07:55,627 --> 00:08:00,877
their customers, so more brain power
more strategic thinking rather than
134
00:08:00,977 --> 00:08:02,687
staying on the low value-added services.
135
00:08:03,397 --> 00:08:05,237
That's really the name of
the game for them as well.
136
00:08:05,337 --> 00:08:09,677
There's gonna be a shift between the
ones adopting it having structured
137
00:08:09,697 --> 00:08:14,997
the data, having structured a strategy
on AI enablement and also potentially
138
00:08:15,056 --> 00:08:18,256
moving to agentic versus the one
which are gonna lag behind, except
139
00:08:18,316 --> 00:08:22,166
if they are very specialized, like
in litigation but otherwise, it's
140
00:08:22,186 --> 00:08:26,666
gonna be a bit challenging for some
entities if they don't adopt AI quickly.
141
00:08:28,526 --> 00:08:34,986
What advice would you share for corporate
legal and law firm leaders, given your
142
00:08:35,046 --> 00:08:41,686
current role but also your vast experience
in technology, in terms of adopting
143
00:08:41,866 --> 00:08:45,287
AI technology in their IP practices?
144
00:08:47,722 --> 00:08:48,882
This wave is incredible.
145
00:08:49,402 --> 00:08:50,802
I've never seen a wave like this.
146
00:08:51,222 --> 00:08:58,939
And I'm now into tech for 20 years, so
I've seen digital IoT crypto, Metaverse,
147
00:08:58,999 --> 00:09:04,319
but this one is a true one, and the
speed and the scale of it is crazy.
148
00:09:04,859 --> 00:09:06,069
It's absolutely insane.
149
00:09:07,089 --> 00:09:09,239
It's exciting, but also a bit frightening.
150
00:09:09,289 --> 00:09:12,199
This AI adoption doesn't come naturally.
151
00:09:12,259 --> 00:09:15,959
Some humans love to change,
some others hate to change.
152
00:09:16,069 --> 00:09:21,539
This human AI adoption needs to be
promoted, needs to be sold , or you need
153
00:09:21,549 --> 00:09:26,539
to convince your people to get into AI
and, of course, put the relevant budgets
154
00:09:26,599 --> 00:09:28,379
to convince your teams to move into AI.
155
00:09:28,499 --> 00:09:32,649
It's an investment at the beginning, and
it's also a lead by example investment.
156
00:09:32,719 --> 00:09:35,939
Within a corporation or law
firms, it also come from the top.
157
00:09:36,028 --> 00:09:38,808
It's not just saying, "We're gonna
do AI," and nobody cares about it.
158
00:09:38,848 --> 00:09:39,768
We launch one program.
159
00:09:39,808 --> 00:09:44,958
It's really about leading the change from
the top as well and giving the means to
160
00:09:44,998 --> 00:09:46,678
people to embrace it and train again.
161
00:09:47,308 --> 00:09:47,658
That's one.
162
00:09:47,748 --> 00:09:49,908
Then second, there are
a few investments to do.
163
00:09:49,978 --> 00:09:52,408
Infrastructure, stack architecture.
164
00:09:53,038 --> 00:09:57,278
The tech stack is critical for
running AI because you need to
165
00:09:57,318 --> 00:09:59,158
have robust tech stack as well.
166
00:09:59,898 --> 00:10:03,438
In terms of system integration, you
need to be able to monitor the cost of
167
00:10:03,818 --> 00:10:06,958
what you're running because it can be
really expensive if you don't monitor.
168
00:10:07,458 --> 00:10:09,748
And your data layers, where are they?
169
00:10:09,888 --> 00:10:12,268
Have you structured your data
in a sense where you can run
170
00:10:12,308 --> 00:10:13,948
your LLMs or your AI on it?
171
00:10:14,248 --> 00:10:18,448
Which is really a critical piece
of it when it comes to, having
172
00:10:18,468 --> 00:10:20,858
a full, a powerful usage of AI.
173
00:10:20,958 --> 00:10:22,748
And then, it's also the way you see it.
174
00:10:23,008 --> 00:10:28,348
Is it just an internal user workspace
or is it also a customer thing?
175
00:10:28,548 --> 00:10:32,208
And how do you deploy that if you want
to give access to some of your customers
176
00:10:32,248 --> 00:10:33,538
depending on the business you run.
177
00:10:33,598 --> 00:10:40,398
Security as well, but this one is
often a bit too, simplified and also
178
00:10:40,518 --> 00:10:44,558
often a bit forgotten when we want to
launch adoption and massive adoption.
179
00:10:45,338 --> 00:10:46,628
In fact it's a critical one.
180
00:10:48,541 --> 00:10:53,382
How can these teams mitigate
security, confidentiality, accuracy,
181
00:10:54,201 --> 00:10:58,601
or quality risks and concerns?
182
00:11:00,912 --> 00:11:03,372
We have developed AI on
top of LLMs we're using.
183
00:11:03,432 --> 00:11:04,632
So we are agnostic to LLMs.
184
00:11:04,682 --> 00:11:07,202
We're testing all the LLMs all
the time, and we're putting our
185
00:11:07,272 --> 00:11:10,152
AI on top of that is forcing the
LLM to work in a certain manner.
186
00:11:11,382 --> 00:11:13,822
We're trying to be more
conservative in the approach.
187
00:11:13,982 --> 00:11:18,502
We don't want to be over, aggressive,
especially when it comes to IP, and
188
00:11:18,642 --> 00:11:23,733
patent and the way we see litigations,
the way we see laws as well.
189
00:11:23,853 --> 00:11:25,393
We have this in mind all the time.
190
00:11:25,923 --> 00:11:30,803
What we're saying to our customers
is, you keep the control of what
191
00:11:30,863 --> 00:11:33,583
you're sending to our customers, so
you need to review what you're doing.
192
00:11:34,203 --> 00:11:37,963
You need to control the outcome of
our tools, and you need to apply
193
00:11:38,043 --> 00:11:40,383
your own human critical sense on it.
194
00:11:40,523 --> 00:11:41,763
You should not just rely on it.
195
00:11:41,833 --> 00:11:45,613
And we've seen some problems in the IP
space where people are getting lazy and
196
00:11:45,663 --> 00:11:47,983
just sending the output of the work.
197
00:11:48,053 --> 00:11:52,573
It's a critical piece in terms of
quality first, on data, I can only
198
00:11:52,613 --> 00:11:57,073
advise people to think about that data
architecture and how they structure
199
00:11:57,093 --> 00:12:00,213
the data, because that's the name
of the game when it comes to AI.
200
00:12:01,353 --> 00:12:04,513
Process mapping is the same as
well, because when you wanna apply
201
00:12:04,633 --> 00:12:08,733
AI into internal processes, or even
external processes, you need to
202
00:12:08,793 --> 00:12:10,823
have clear process mapping as well.
203
00:12:10,853 --> 00:12:12,793
That's a bit of preliminary
work you need to do.
204
00:12:13,923 --> 00:12:17,833
For security, it's making sure
you have strong, IT security,
205
00:12:17,953 --> 00:12:19,373
and data privacy as well.
206
00:12:19,613 --> 00:12:22,573
That's the second element especially,
it's a bit stronger in Europe, as we're
207
00:12:22,603 --> 00:12:25,883
the specialist and the champions of
regulation, but at least we're ready
208
00:12:26,143 --> 00:12:28,563
for it now thanks to our dear GDPR.
209
00:12:29,043 --> 00:12:34,103
We stick by the rules and that's also
clearly vis-à-vis our customers, it's also
210
00:12:34,143 --> 00:12:36,202
a kind of of course an important argument.
211
00:12:36,302 --> 00:12:41,212
But it's critical as well, especially when
you deal with confidential and critical
212
00:12:41,272 --> 00:12:43,712
information from from your customer data.
213
00:12:43,802 --> 00:12:45,062
So that's really important as well.
214
00:12:46,102 --> 00:12:50,722
How do you see AI affecting the IP
legal sector over the long term?
215
00:12:52,142 --> 00:12:54,182
AI is a huge success.
216
00:12:54,242 --> 00:12:59,012
It's an incredible wave, but there's also
a bit of communication, marketing hype.
217
00:12:59,052 --> 00:13:04,192
On our side, we're trying to stay
agnostic, and not relying heavily on one.
218
00:13:04,312 --> 00:13:07,452
We're using preference two today
because they are giving us the best
219
00:13:07,592 --> 00:13:13,232
results vis-a-vis our own developed
AI and also on our agentic structure.
220
00:13:13,292 --> 00:13:16,072
But we're testing all the AI all the
time and the LLM all the time, and
221
00:13:16,122 --> 00:13:20,542
we think that tomorrow AI is gonna
clearly change the way we work.
222
00:13:20,592 --> 00:13:23,232
It's already the case, but it's
gonna accelerate more and more.
223
00:13:24,092 --> 00:13:27,272
So there's gonna be a shift in
the way people work, but also the
224
00:13:27,372 --> 00:13:28,812
structure of a company as well.
225
00:13:28,912 --> 00:13:32,602
Low administrative tasks are
gonna be more and more executed
226
00:13:32,622 --> 00:13:34,542
automatically with a human control.
227
00:13:35,302 --> 00:13:38,842
But it's not at all the same when
you know that today you can reverse
228
00:13:38,882 --> 00:13:42,372
engineer a software not a huge
complex software, but you can reverse
229
00:13:42,532 --> 00:13:44,182
engineer a software in two days.
230
00:13:44,282 --> 00:13:45,222
That's the reality of it.
231
00:13:45,282 --> 00:13:49,862
If you take a few of the big LLMs, and
I'm not gonna do advertising for any
232
00:13:49,902 --> 00:13:51,862
of them, but that's the reality of it.
233
00:13:51,922 --> 00:13:54,542
So it costs a lot of tokens, of course.
234
00:13:55,152 --> 00:13:55,912
It's not for free.
235
00:13:56,652 --> 00:13:59,942
But but it's the reality
of the power of this tool.
236
00:14:00,042 --> 00:14:03,302
It's gonna change massively
some of the administrative work.
237
00:14:03,382 --> 00:14:06,902
We see a shift also from software
developers to AI developers, And that's
238
00:14:06,942 --> 00:14:10,162
also the scarcity of the resource
you face, because you don't find as
239
00:14:10,202 --> 00:14:14,182
many AI developers and AI engineers
than software developers today.
240
00:14:14,932 --> 00:14:17,662
That's why I'm saying to my teams,
if you're not willing to change, do
241
00:14:17,702 --> 00:14:21,352
it for yourself, don't do it for the
company, but do it for yourself because
242
00:14:21,472 --> 00:14:25,992
your capacity to be employed tomorrow
is gonna be related to this as well.
243
00:14:26,132 --> 00:14:31,422
That's a critical aspect for you, not
only for our company because it's a race.
244
00:14:31,482 --> 00:14:33,342
It's gonna be a bit of a
marathon, but it's a race.
245
00:14:33,712 --> 00:14:39,392
And being agnostic it's one of the key
aspects, being ready for transformation.
246
00:14:39,892 --> 00:14:42,853
So having our people adopting
this wave is critical.
247
00:14:43,552 --> 00:14:46,873
And the future nobody really knows
except that we think that agents are
248
00:14:46,892 --> 00:14:50,212
gonna be really there for some time
with the right human control, of course.
249
00:14:52,073 --> 00:14:58,172
This is Ari Kaplan speaking with
Frédéric Beylier, the CEO of Questel,
250
00:14:58,253 --> 00:15:01,212
a provider of intellectual property
solutions, including AI-powered
251
00:15:01,372 --> 00:15:04,032
modular software and expert services.
252
00:15:04,732 --> 00:15:06,792
Fred, thanks so very much,
and I'm wishing you the very
253
00:15:06,832 --> 00:15:08,372
best of luck in your new role.
254
00:15:09,452 --> 00:15:10,172
Thank you, Ari.
255
00:15:10,232 --> 00:15:11,273
Thank you for your time.
256
00:15:11,332 --> 00:15:12,152
Thank you for the chat.
257
00:15:13,722 --> 00:15:17,822
Thank you for listening to the
Reinventing Professionals podcast.
258
00:15:17,862 --> 00:15:24,362
Visit reinventingprofessionals.com or
arikaplanadvisors.com to learn more.
00:00:01,357 --> 00:00:05,897
Welcome to Reinventing Professionals,
a podcast hosted by industry analyst
2
00:00:05,977 --> 00:00:10,777
Ari Kaplan, which shares ideas,
guidance, and perspectives from market
3
00:00:10,837 --> 00:00:15,118
leaders shaping the next generation
of legal and professional services.
4
00:00:19,068 --> 00:00:24,628
This is Ari Kaplan, and I'm speaking
today with Frédéric Beylier, the CEO
5
00:00:24,708 --> 00:00:29,108
of Questel, a provider of intellectual
property solutions, including AI-powered
6
00:00:29,669 --> 00:00:32,608
modular software and expert services.
7
00:00:33,088 --> 00:00:33,448
Hi, Fred.
8
00:00:33,488 --> 00:00:33,869
How are you?
9
00:00:34,648 --> 00:00:35,248
Hi, Ari.
10
00:00:35,328 --> 00:00:36,228
Thanks for having me.
11
00:00:36,408 --> 00:00:37,528
Great to speak to you today.
12
00:00:38,108 --> 00:00:39,428
It's my privilege to speak with you.
13
00:00:39,469 --> 00:00:43,228
So tell us about your background
and your role at Questel.
14
00:00:44,548 --> 00:00:47,529
I've been running companies in the
tech space for the past 20 years now.
15
00:00:48,388 --> 00:00:51,348
Initially in the security
technology space.
16
00:00:51,439 --> 00:00:56,289
Then I moved also to the mobility tech
mobility space, the urban mobility.
17
00:00:57,149 --> 00:00:58,970
And now I'm in IP and I love it.
18
00:00:58,990 --> 00:01:00,890
There are some similarities, of course.
19
00:01:01,030 --> 00:01:04,449
Most of the time I'm getting in industry
about, there's a kind of transformation
20
00:01:04,490 --> 00:01:10,609
ongoing, so at the time of the security
technology was digital, and IoT, and
21
00:01:10,709 --> 00:01:15,430
now it's AI for IP, so really exciting
and trying to build the industry leaders
22
00:01:15,451 --> 00:01:19,121
in, in their space, of course, that's
the intent is really to try to build
23
00:01:19,221 --> 00:01:20,601
industry leaders in the long run.
24
00:01:20,691 --> 00:01:23,691
How is AI changing the IP sector?
25
00:01:25,378 --> 00:01:26,308
It has changed already.
26
00:01:26,458 --> 00:01:31,098
We did a report on AI adoption
and IP, and fifty percent of the
27
00:01:31,138 --> 00:01:34,878
respondents are declaring that
they're using AI at least once a day.
28
00:01:35,228 --> 00:01:38,028
Of course it's AI assistance,
so it's just the start of it.
29
00:01:39,128 --> 00:01:43,018
But we believe that in enterprise
as a whole and not IP, but
30
00:01:43,038 --> 00:01:46,468
enterprise the rate is more around
twenty to twenty-five percent.
31
00:01:46,548 --> 00:01:50,388
So having respondents in the IP
space answering fifty-plus percent
32
00:01:50,968 --> 00:01:53,548
at least once a day, it's not bad.
33
00:01:54,108 --> 00:01:55,168
It's pretty advanced.
34
00:01:55,558 --> 00:01:59,578
It's about AI assistance, the way you
work, the way you get a summary, the
35
00:01:59,638 --> 00:02:02,248
way you get an an opinion or you prompt.
36
00:02:02,928 --> 00:02:06,858
It can summarize a lot of data
information, and some of the tools are
37
00:02:06,878 --> 00:02:11,178
already providing this, Which are really
AI native tools like our tool in Qthena.
38
00:02:11,238 --> 00:02:14,128
Tomorrow we're gonna see more and
more agents coming up and that's what
39
00:02:14,168 --> 00:02:18,628
it's about now and helping people
executing tasks, and multiplying and
40
00:02:18,658 --> 00:02:21,848
leveraging the power of agents to
do the work on behalf of the people.
41
00:02:21,938 --> 00:02:26,938
What are the emerging challenges
that brand owners and innovators
42
00:02:26,978 --> 00:02:28,758
are facing in the current climate?
43
00:02:30,328 --> 00:02:35,488
Multiple because, of course, when we come
to innovations it's always more complex.
44
00:02:35,548 --> 00:02:36,608
You need to handle AI.
45
00:02:36,688 --> 00:02:41,258
So there are native AI players that
are used to this, but you also have
46
00:02:41,358 --> 00:02:46,538
a complexity in the way you to some
extent patent your ideas as well.
47
00:02:46,588 --> 00:02:51,218
Now it's AI open source technology and
you need to be really strong on the way
48
00:02:51,378 --> 00:02:53,358
you patent your technologies as well.
49
00:02:53,408 --> 00:02:58,648
Today there are key challenges which
are, having a homogenous adoption
50
00:02:58,688 --> 00:03:00,287
of AI, around the innovators.
51
00:03:00,388 --> 00:03:03,318
And at the same time it's an adoption
and you want a change to happen,
52
00:03:03,398 --> 00:03:06,289
but you have key constraints,
do you have proprietary data?
53
00:03:06,549 --> 00:03:07,289
Who's the data?
54
00:03:08,089 --> 00:03:09,309
Who's belonging the data?
55
00:03:09,849 --> 00:03:12,109
Do you have security at
the right level as well?
56
00:03:13,049 --> 00:03:16,009
Do you have the right ecosystem
around you in terms of integration?
57
00:03:16,368 --> 00:03:21,988
So there are a few technical complexities
when it comes to challenges of owners and
58
00:03:22,069 --> 00:03:26,368
innovators, brand owners and innovators
when it comes to AI, which are pretty big
59
00:03:26,428 --> 00:03:31,549
because you can jump into AI, but you're
also in the open world, I would say.
60
00:03:31,658 --> 00:03:35,438
And using classical LLM, you put
part of your innovation, part of your
61
00:03:36,459 --> 00:03:40,668
secret sauce, in the open world, and
that's also some critical aspects.
62
00:03:40,719 --> 00:03:43,559
And the way you scale AI is not
gonna be the same if your structure
63
00:03:43,600 --> 00:03:45,979
in term of data before that or not.
64
00:03:45,999 --> 00:03:48,100
It's really different the
way you do it as well.
65
00:03:48,179 --> 00:03:49,899
It's the how that is important.
66
00:03:49,989 --> 00:03:55,369
How is Questel addressing the risks
and opportunities associated with AI?
67
00:03:56,009 --> 00:03:59,529
We see 200% opportunities more
than risk to be honest with you.
68
00:04:00,110 --> 00:04:03,989
We see it as an opportunity because
it's an accelerator for innovation.
69
00:04:04,979 --> 00:04:06,850
We have a few AI-native products.
70
00:04:07,629 --> 00:04:12,189
Some of our more software-based product
are also now embedded with AI as well.
71
00:04:12,210 --> 00:04:12,771
We're using it.
72
00:04:12,811 --> 00:04:18,199
We have developed our own AI LLM that
we embed into our products as well.
73
00:04:18,699 --> 00:04:19,799
That's on the product side.
74
00:04:19,859 --> 00:04:21,099
And we're moving to agentic.
75
00:04:21,150 --> 00:04:25,531
We're now launching some agentic solution,
especially in our Qthena, which is really
76
00:04:25,590 --> 00:04:29,950
a kind of AI-native cockpit helping,
patent attorneys or trademark attorneys
77
00:04:29,970 --> 00:04:36,261
to draft patents, summarize patents- do
checks get all the public information.
78
00:04:36,361 --> 00:04:39,701
They can even add their proprietary
information inside it, so it's really
79
00:04:39,801 --> 00:04:43,501
a kind of cockpit for the patent
attorney to gain approximately 40%
80
00:04:43,521 --> 00:04:44,701
of their time, so that's massive.
81
00:04:45,441 --> 00:04:48,571
For us it's really an opportunity from
a business development standpoint.
82
00:04:49,271 --> 00:04:53,451
It's an opportunity as well from
an efficiency, internal efficiency,
83
00:04:53,531 --> 00:04:58,171
because the more we adopt AI internally,
the more we map our processes and
84
00:04:58,251 --> 00:05:01,591
we use AI, the more we automate our
processes, the more we are efficient.
85
00:05:02,471 --> 00:05:05,141
So we can spare time, on
more value-added task.
86
00:05:05,251 --> 00:05:06,411
Now, it's also a risk.
87
00:05:07,161 --> 00:05:10,530
The risk is more on the big LLMs
because everybody thinks they
88
00:05:10,621 --> 00:05:13,341
can do everything with the big
LLM, which is not exactly true.
89
00:05:13,761 --> 00:05:18,631
The reality of it is depending on
the depth of the work you need to do,
90
00:05:18,791 --> 00:05:21,331
you want to do the data is critical.
91
00:05:21,431 --> 00:05:25,311
So you want to have access at the same
data in an LLM than in other areas, and
92
00:05:25,411 --> 00:05:30,741
es-especially when you're deep into a
vertical like IP the data you can provide
93
00:05:30,791 --> 00:05:35,461
is much more precise and accurate and
as well qualified than your current LLM.
94
00:05:36,381 --> 00:05:38,301
And also, LLMs have
hallucinations as well.
95
00:05:38,601 --> 00:05:39,721
So you need to control that.
96
00:05:39,781 --> 00:05:42,441
You need to be on top
of the results you get.
97
00:05:42,541 --> 00:05:44,541
We see LLMs a-as a competitor.
98
00:05:44,621 --> 00:05:48,821
It's a key enabler for us as well
because we're using them and we combine
99
00:05:48,831 --> 00:05:51,081
them with our own AI we've developed.
100
00:05:51,941 --> 00:05:56,071
But also it's a competition for us
because we look at what they do and how
101
00:05:56,101 --> 00:06:03,237
they disrupt potentially the usage. And
they have for them huge fundraising,
102
00:06:03,317 --> 00:06:05,477
they have for them huge marketing spend.
103
00:06:06,357 --> 00:06:10,537
Now what we have for us is
proprietary data in IP, so
104
00:06:10,697 --> 00:06:12,077
really qualified and enriched.
105
00:06:12,937 --> 00:06:16,317
We have for us a deep
vertical functional knowledge.
106
00:06:17,177 --> 00:06:20,816
We are also, executing some
critical tasks for our customers.
107
00:06:21,197 --> 00:06:23,677
We are paying annuities
for patent annuities.
108
00:06:23,737 --> 00:06:25,917
We are renewing marks on
behalf of our customers.
109
00:06:26,507 --> 00:06:30,647
We're doing a lot of depth of
services, but also the functional
110
00:06:31,447 --> 00:06:35,447
know-how is way deeper than, of
course, what you can find in LLM.
111
00:06:35,507 --> 00:06:38,807
So that's what we have for us,
and we need to go deeper and
112
00:06:38,847 --> 00:06:42,987
deeper into our expertise, of
course, to, to be differentiated
113
00:06:43,047 --> 00:06:44,566
vis-à-vis the LLM for our customers.
114
00:06:45,987 --> 00:06:48,987
You mentioned that Questel recently
released benchmarking research
115
00:06:49,027 --> 00:06:52,147
on the adoption- Yeah … of
AI tools across the IP sector.
116
00:06:52,307 --> 00:06:59,687
What did you identify as the main
use cases of AI for IP, and where
117
00:07:00,007 --> 00:07:03,807
are IP professionals realizing
the greatest benefits from AI?
118
00:07:05,087 --> 00:07:08,307
It's really about AI assistance
and how do I gain time today.
119
00:07:09,287 --> 00:07:12,147
And that's most of the
outcome we see today.
120
00:07:12,187 --> 00:07:15,267
Everybody now acknowledges that
AI is gonna change the way I work.
121
00:07:15,327 --> 00:07:18,307
It's seventy-four percent now
acknowledging this across the
122
00:07:18,327 --> 00:07:20,087
IP space, so that's significant.
123
00:07:21,097 --> 00:07:25,297
It's a matter of efficiency being
augmented, I would say, like more
124
00:07:25,317 --> 00:07:26,657
productive and more efficient.
125
00:07:27,737 --> 00:07:31,297
Tomorrow everybody is not seeing
the agentic piece because not
126
00:07:31,357 --> 00:07:35,197
everybody's advanced as as some of
the players, and especially us, we
127
00:07:35,217 --> 00:07:38,737
are tech providers, so we see the
agentic, we develop it, we launch it.
128
00:07:38,827 --> 00:07:42,397
But it's gonna also change the value
chain in IP, I would say, because
129
00:07:42,477 --> 00:07:46,817
some basic services that you used
to call a third-party provider, you
130
00:07:46,877 --> 00:07:49,867
can do it through agents now, so
it's really different in the way
131
00:07:49,887 --> 00:07:50,947
it's gonna change the value chain.
132
00:07:51,037 --> 00:07:55,597
For anybody in the IP space, lawyers,
they need to bring more added value to
133
00:07:55,627 --> 00:08:00,877
their customers, so more brain power
more strategic thinking rather than
134
00:08:00,977 --> 00:08:02,687
staying on the low value-added services.
135
00:08:03,397 --> 00:08:05,237
That's really the name of
the game for them as well.
136
00:08:05,337 --> 00:08:09,677
There's gonna be a shift between the
ones adopting it having structured
137
00:08:09,697 --> 00:08:14,997
the data, having structured a strategy
on AI enablement and also potentially
138
00:08:15,056 --> 00:08:18,256
moving to agentic versus the one
which are gonna lag behind, except
139
00:08:18,316 --> 00:08:22,166
if they are very specialized, like
in litigation but otherwise, it's
140
00:08:22,186 --> 00:08:26,666
gonna be a bit challenging for some
entities if they don't adopt AI quickly.
141
00:08:28,526 --> 00:08:34,986
What advice would you share for corporate
legal and law firm leaders, given your
142
00:08:35,046 --> 00:08:41,686
current role but also your vast experience
in technology, in terms of adopting
143
00:08:41,866 --> 00:08:45,287
AI technology in their IP practices?
144
00:08:47,722 --> 00:08:48,882
This wave is incredible.
145
00:08:49,402 --> 00:08:50,802
I've never seen a wave like this.
146
00:08:51,222 --> 00:08:58,939
And I'm now into tech for 20 years, so
I've seen digital IoT crypto, Metaverse,
147
00:08:58,999 --> 00:09:04,319
but this one is a true one, and the
speed and the scale of it is crazy.
148
00:09:04,859 --> 00:09:06,069
It's absolutely insane.
149
00:09:07,089 --> 00:09:09,239
It's exciting, but also a bit frightening.
150
00:09:09,289 --> 00:09:12,199
This AI adoption doesn't come naturally.
151
00:09:12,259 --> 00:09:15,959
Some humans love to change,
some others hate to change.
152
00:09:16,069 --> 00:09:21,539
This human AI adoption needs to be
promoted, needs to be sold , or you need
153
00:09:21,549 --> 00:09:26,539
to convince your people to get into AI
and, of course, put the relevant budgets
154
00:09:26,599 --> 00:09:28,379
to convince your teams to move into AI.
155
00:09:28,499 --> 00:09:32,649
It's an investment at the beginning, and
it's also a lead by example investment.
156
00:09:32,719 --> 00:09:35,939
Within a corporation or law
firms, it also come from the top.
157
00:09:36,028 --> 00:09:38,808
It's not just saying, "We're gonna
do AI," and nobody cares about it.
158
00:09:38,848 --> 00:09:39,768
We launch one program.
159
00:09:39,808 --> 00:09:44,958
It's really about leading the change from
the top as well and giving the means to
160
00:09:44,998 --> 00:09:46,678
people to embrace it and train again.
161
00:09:47,308 --> 00:09:47,658
That's one.
162
00:09:47,748 --> 00:09:49,908
Then second, there are
a few investments to do.
163
00:09:49,978 --> 00:09:52,408
Infrastructure, stack architecture.
164
00:09:53,038 --> 00:09:57,278
The tech stack is critical for
running AI because you need to
165
00:09:57,318 --> 00:09:59,158
have robust tech stack as well.
166
00:09:59,898 --> 00:10:03,438
In terms of system integration, you
need to be able to monitor the cost of
167
00:10:03,818 --> 00:10:06,958
what you're running because it can be
really expensive if you don't monitor.
168
00:10:07,458 --> 00:10:09,748
And your data layers, where are they?
169
00:10:09,888 --> 00:10:12,268
Have you structured your data
in a sense where you can run
170
00:10:12,308 --> 00:10:13,948
your LLMs or your AI on it?
171
00:10:14,248 --> 00:10:18,448
Which is really a critical piece
of it when it comes to, having
172
00:10:18,468 --> 00:10:20,858
a full, a powerful usage of AI.
173
00:10:20,958 --> 00:10:22,748
And then, it's also the way you see it.
174
00:10:23,008 --> 00:10:28,348
Is it just an internal user workspace
or is it also a customer thing?
175
00:10:28,548 --> 00:10:32,208
And how do you deploy that if you want
to give access to some of your customers
176
00:10:32,248 --> 00:10:33,538
depending on the business you run.
177
00:10:33,598 --> 00:10:40,398
Security as well, but this one is
often a bit too, simplified and also
178
00:10:40,518 --> 00:10:44,558
often a bit forgotten when we want to
launch adoption and massive adoption.
179
00:10:45,338 --> 00:10:46,628
In fact it's a critical one.
180
00:10:48,541 --> 00:10:53,382
How can these teams mitigate
security, confidentiality, accuracy,
181
00:10:54,201 --> 00:10:58,601
or quality risks and concerns?
182
00:11:00,912 --> 00:11:03,372
We have developed AI on
top of LLMs we're using.
183
00:11:03,432 --> 00:11:04,632
So we are agnostic to LLMs.
184
00:11:04,682 --> 00:11:07,202
We're testing all the LLMs all
the time, and we're putting our
185
00:11:07,272 --> 00:11:10,152
AI on top of that is forcing the
LLM to work in a certain manner.
186
00:11:11,382 --> 00:11:13,822
We're trying to be more
conservative in the approach.
187
00:11:13,982 --> 00:11:18,502
We don't want to be over, aggressive,
especially when it comes to IP, and
188
00:11:18,642 --> 00:11:23,733
patent and the way we see litigations,
the way we see laws as well.
189
00:11:23,853 --> 00:11:25,393
We have this in mind all the time.
190
00:11:25,923 --> 00:11:30,803
What we're saying to our customers
is, you keep the control of what
191
00:11:30,863 --> 00:11:33,583
you're sending to our customers, so
you need to review what you're doing.
192
00:11:34,203 --> 00:11:37,963
You need to control the outcome of
our tools, and you need to apply
193
00:11:38,043 --> 00:11:40,383
your own human critical sense on it.
194
00:11:40,523 --> 00:11:41,763
You should not just rely on it.
195
00:11:41,833 --> 00:11:45,613
And we've seen some problems in the IP
space where people are getting lazy and
196
00:11:45,663 --> 00:11:47,983
just sending the output of the work.
197
00:11:48,053 --> 00:11:52,573
It's a critical piece in terms of
quality first, on data, I can only
198
00:11:52,613 --> 00:11:57,073
advise people to think about that data
architecture and how they structure
199
00:11:57,093 --> 00:12:00,213
the data, because that's the name
of the game when it comes to AI.
200
00:12:01,353 --> 00:12:04,513
Process mapping is the same as
well, because when you wanna apply
201
00:12:04,633 --> 00:12:08,733
AI into internal processes, or even
external processes, you need to
202
00:12:08,793 --> 00:12:10,823
have clear process mapping as well.
203
00:12:10,853 --> 00:12:12,793
That's a bit of preliminary
work you need to do.
204
00:12:13,923 --> 00:12:17,833
For security, it's making sure
you have strong, IT security,
205
00:12:17,953 --> 00:12:19,373
and data privacy as well.
206
00:12:19,613 --> 00:12:22,573
That's the second element especially,
it's a bit stronger in Europe, as we're
207
00:12:22,603 --> 00:12:25,883
the specialist and the champions of
regulation, but at least we're ready
208
00:12:26,143 --> 00:12:28,563
for it now thanks to our dear GDPR.
209
00:12:29,043 --> 00:12:34,103
We stick by the rules and that's also
clearly vis-à-vis our customers, it's also
210
00:12:34,143 --> 00:12:36,202
a kind of of course an important argument.
211
00:12:36,302 --> 00:12:41,212
But it's critical as well, especially when
you deal with confidential and critical
212
00:12:41,272 --> 00:12:43,712
information from from your customer data.
213
00:12:43,802 --> 00:12:45,062
So that's really important as well.
214
00:12:46,102 --> 00:12:50,722
How do you see AI affecting the IP
legal sector over the long term?
215
00:12:52,142 --> 00:12:54,182
AI is a huge success.
216
00:12:54,242 --> 00:12:59,012
It's an incredible wave, but there's also
a bit of communication, marketing hype.
217
00:12:59,052 --> 00:13:04,192
On our side, we're trying to stay
agnostic, and not relying heavily on one.
218
00:13:04,312 --> 00:13:07,452
We're using preference two today
because they are giving us the best
219
00:13:07,592 --> 00:13:13,232
results vis-a-vis our own developed
AI and also on our agentic structure.
220
00:13:13,292 --> 00:13:16,072
But we're testing all the AI all the
time and the LLM all the time, and
221
00:13:16,122 --> 00:13:20,542
we think that tomorrow AI is gonna
clearly change the way we work.
222
00:13:20,592 --> 00:13:23,232
It's already the case, but it's
gonna accelerate more and more.
223
00:13:24,092 --> 00:13:27,272
So there's gonna be a shift in
the way people work, but also the
224
00:13:27,372 --> 00:13:28,812
structure of a company as well.
225
00:13:28,912 --> 00:13:32,602
Low administrative tasks are
gonna be more and more executed
226
00:13:32,622 --> 00:13:34,542
automatically with a human control.
227
00:13:35,302 --> 00:13:38,842
But it's not at all the same when
you know that today you can reverse
228
00:13:38,882 --> 00:13:42,372
engineer a software not a huge
complex software, but you can reverse
229
00:13:42,532 --> 00:13:44,182
engineer a software in two days.
230
00:13:44,282 --> 00:13:45,222
That's the reality of it.
231
00:13:45,282 --> 00:13:49,862
If you take a few of the big LLMs, and
I'm not gonna do advertising for any
232
00:13:49,902 --> 00:13:51,862
of them, but that's the reality of it.
233
00:13:51,922 --> 00:13:54,542
So it costs a lot of tokens, of course.
234
00:13:55,152 --> 00:13:55,912
It's not for free.
235
00:13:56,652 --> 00:13:59,942
But but it's the reality
of the power of this tool.
236
00:14:00,042 --> 00:14:03,302
It's gonna change massively
some of the administrative work.
237
00:14:03,382 --> 00:14:06,902
We see a shift also from software
developers to AI developers, And that's
238
00:14:06,942 --> 00:14:10,162
also the scarcity of the resource
you face, because you don't find as
239
00:14:10,202 --> 00:14:14,182
many AI developers and AI engineers
than software developers today.
240
00:14:14,932 --> 00:14:17,662
That's why I'm saying to my teams,
if you're not willing to change, do
241
00:14:17,702 --> 00:14:21,352
it for yourself, don't do it for the
company, but do it for yourself because
242
00:14:21,472 --> 00:14:25,992
your capacity to be employed tomorrow
is gonna be related to this as well.
243
00:14:26,132 --> 00:14:31,422
That's a critical aspect for you, not
only for our company because it's a race.
244
00:14:31,482 --> 00:14:33,342
It's gonna be a bit of a
marathon, but it's a race.
245
00:14:33,712 --> 00:14:39,392
And being agnostic it's one of the key
aspects, being ready for transformation.
246
00:14:39,892 --> 00:14:42,853
So having our people adopting
this wave is critical.
247
00:14:43,552 --> 00:14:46,873
And the future nobody really knows
except that we think that agents are
248
00:14:46,892 --> 00:14:50,212
gonna be really there for some time
with the right human control, of course.
249
00:14:52,073 --> 00:14:58,172
This is Ari Kaplan speaking with
Frédéric Beylier, the CEO of Questel,
250
00:14:58,253 --> 00:15:01,212
a provider of intellectual property
solutions, including AI-powered
251
00:15:01,372 --> 00:15:04,032
modular software and expert services.
252
00:15:04,732 --> 00:15:06,792
Fred, thanks so very much,
and I'm wishing you the very
253
00:15:06,832 --> 00:15:08,372
best of luck in your new role.
254
00:15:09,452 --> 00:15:10,172
Thank you, Ari.
255
00:15:10,232 --> 00:15:11,273
Thank you for your time.
256
00:15:11,332 --> 00:15:12,152
Thank you for the chat.
257
00:15:13,722 --> 00:15:17,822
Thank you for listening to the
Reinventing Professionals podcast.
258
00:15:17,862 --> 00:15:24,362
Visit reinventingprofessionals.com or
arikaplanadvisors.com to learn more.
