How to Build an AI Strategy That Scales Beyond the Pilot
AI could influence 90% of jobs. David Lloyd says AI literacy is still only around 38%.
That gap is where many enterprise AI strategies begin to break.
In the latest HR Leaders Podcast episode, I spoke with David Lloyd, SVP Platform Engineering & Chief AI Officer at Dayforce. He explains why HR cannot wait until the end of an AI transformation to address skills, trust and organisational change.
Most organisations do not lack AI ideas. They lack trusted data, safe environments and a clear process for deciding which ideas deserve to scale.
And even when an experiment works, a harder question remains:
Can it scale from one employee to 5,000?
5 things you’ll learn from this episode:
Why three to six months is too long to judge an AI pilot
The three questions that decide whether an AI idea moves forward
What is really driving the rise of shadow AI
Why adding AI across a fragmented HR stack creates more chaos
How a three-person governance team can approve ideas in under five days
🧠 Your learning platform proves who finished. Can it prove who is ready to perform?
Most training takes weeks to build, gives everyone the same content and stops at completion. By the time it launches, the business may have already moved on.
Fluide starts with the outcome, adapts learning to each person and uses realistic practice to show who is improving, who needs reinforcement and who is ready to perform.
Learn how Fluide turns a skill gap into live adaptive learning today and see performance move by Friday
1
00:00:00.340 --> 00:00:04.320
If you're not driving quantifiable value or real value out of the process,
2
00:00:04.860 --> 00:00:05.340
stop it.
3
00:00:06.220 --> 00:00:08.480
It's not going to get necessarily better over time.
4
00:00:08.600 --> 00:00:11.900
So be harsh in stopping those things that aren't working.
5
00:00:12.380 --> 00:00:13.720
And then those that are working,
6
00:00:15.040 --> 00:00:18.480
what you really need to understand is, how do you scale that?
7
00:00:18.540 --> 00:00:21.570
It's one thing when you have an individual who's doing something off the side of
8
00:00:21.620 --> 00:00:23.120
their desk that works well for them.
9
00:00:24.180 --> 00:00:24.480
But
10
00:00:25.480 --> 00:00:29.240
let's say it works really well. How do you scale that out to 4,000 or 5,000 people,
11
00:00:29.960 --> 00:00:33.840
or hundreds? And I don't think we give a lot of thought to that
12
00:00:33.900 --> 00:00:37.640
as well. So we have infrastructures and everything else that aren't really meant to
13
00:00:37.720 --> 00:00:40.160
scale, so we're kind of holding ourselves back.
14
00:00:40.280 --> 00:00:43.800
So I think there's some very basic lessons learned that aren't just about
15
00:00:44.120 --> 00:00:47.280
necessarily AI. I think they've existed in all the different
16
00:00:48.340 --> 00:00:51.640
pieces of technology we've seen come in that are more transformational in an
17
00:00:51.700 --> 00:00:54.740
organization. So I think similar rules apply.
18
00:01:11.520 --> 00:01:13.160
David, welcome to the show, my friend. How are you doing?
19
00:01:14.240 --> 00:01:16.780
I'm doing great, Chris. Great to be here.
20
00:01:17.020 --> 00:01:17.260
It's
21
00:01:18.100 --> 00:01:20.940
been a crazy six months already in 2026.
22
00:01:21.760 --> 00:01:22.380
Are we already
23
00:01:23.440 --> 00:01:25.760
six months... Where did six months go?
24
00:01:26.680 --> 00:01:28.580
I don't know. It's a blank.
25
00:01:29.140 --> 00:01:32.750
Yeah. Do you think that's because of the pace of change right now, like things
26
00:01:32.900 --> 00:01:36.270
moving so fast and we're just kind of overwhelmed, just trying to keep up?
27
00:01:36.340 --> 00:01:38.020
Do you think that's kind of a contribution?
28
00:01:38.060 --> 00:01:41.340
I actually do. I think it's a big part of it. Think about it.
29
00:01:41.660 --> 00:01:45.580
It took us, what, 16, 17 years to get to 100
30
00:01:45.600 --> 00:01:47.400
million people using something as
31
00:01:48.240 --> 00:01:49.679
simple as a mobile phone.
32
00:01:49.740 --> 00:01:50.060
Mm.
33
00:01:50.820 --> 00:01:54.060
And arguably, ChatGPT did that in three months.
34
00:01:54.280 --> 00:01:56.260
So that compression, I think that we feel
35
00:01:57.180 --> 00:02:00.600
for the speed at which the technology is not only coming into
36
00:02:01.080 --> 00:02:03.860
our use, but arguably changing. It's not
37
00:02:04.680 --> 00:02:05.160
static.
38
00:02:06.180 --> 00:02:09.830
You walk around with a box you talk on, that was fine for a period of time, but
39
00:02:09.880 --> 00:02:12.570
the change in AI in the last six months-
40
00:02:13.000 --> 00:02:13.090
Yeah
41
00:02:13.360 --> 00:02:16.820
... has been incredible. And so I think we're all feeling that kind of
42
00:02:16.860 --> 00:02:18.000
compression or pressure
43
00:02:18.880 --> 00:02:21.040
based on that constant rate of change.
44
00:02:21.100 --> 00:02:24.960
Yeah. It feels like just when you felt like you've caught up and
45
00:02:25.000 --> 00:02:28.360
you've up-skilled or educated myself, it just
46
00:02:28.460 --> 00:02:29.660
exponentially
47
00:02:30.680 --> 00:02:34.590
changes. You're like, "Oh, I've got to learn around agent automations and
48
00:02:34.620 --> 00:02:37.020
workflows now. I'm not just asking a question."
49
00:02:37.080 --> 00:02:37.190
Mm-hmm.
50
00:02:37.230 --> 00:02:40.580
"I'm now connecting this to my day-to-day life
51
00:02:40.660 --> 00:02:44.180
routine, how it works, and building, orchestrating
52
00:02:44.480 --> 00:02:47.740
agents, not just asking questions." And that's a whole other
53
00:02:47.840 --> 00:02:49.420
skill that we have to-
54
00:02:49.440 --> 00:02:52.280
But then the skill on top of that now is specifications.
55
00:02:52.360 --> 00:02:52.540
So
56
00:02:53.760 --> 00:02:57.600
as I have a big accountability for a large software engineering group, not only
57
00:02:57.660 --> 00:02:57.940
is
58
00:02:58.740 --> 00:03:02.280
that role, but also the chief AI officer, for me, the biggest
59
00:03:02.340 --> 00:03:06.040
challenge is specifications. Everything old is new again.
60
00:03:06.080 --> 00:03:09.700
So now the impact of specifications on how your
61
00:03:09.760 --> 00:03:12.040
agents function is so critical.
62
00:03:12.060 --> 00:03:12.230
Mm-hmm.
63
00:03:12.380 --> 00:03:15.420
So all of these things are just rapidly changing
64
00:03:16.500 --> 00:03:19.400
how we thought of maybe a year or two ago,
65
00:03:19.940 --> 00:03:23.620
prompting. What a neat thing, and now prompting is so passé
66
00:03:23.780 --> 00:03:24.120
and-
67
00:03:24.960 --> 00:03:25.010
Yeah
68
00:03:25.010 --> 00:03:28.340
... so simplistic. So yes, things are moving really quick.
69
00:03:28.820 --> 00:03:29.960
I was saying that to my friend the other day.
70
00:03:30.000 --> 00:03:33.560
I was like, "Prompting was the fastest job to come and go I've ever
71
00:03:33.600 --> 00:03:34.580
seen."
72
00:03:35.580 --> 00:03:37.980
Remember, everyone needed a prompt engineer.
73
00:03:38.000 --> 00:03:41.549
There was all these courses that came out, and now we're like, "Yeah, we don't need
74
00:03:41.560 --> 00:03:45.360
a prompts engineer." The platform's smart enough.
75
00:03:45.480 --> 00:03:49.100
These days, you can type in the most
76
00:03:49.200 --> 00:03:52.480
basic of prompts and actually get a really great response.
77
00:03:53.280 --> 00:03:56.790
Well, that's also how the companies have done a great job of going
78
00:03:57.340 --> 00:04:00.320
with the tokenization part of it. "Oh, great.
79
00:04:00.580 --> 00:04:01.960
My prompt looks pretty tight there.
80
00:04:02.060 --> 00:04:05.160
I think that's only 20 tokens." And in the background, it's now
81
00:04:05.720 --> 00:04:08.940
20,000 tokens that actually gets blown out to actually
82
00:04:09.640 --> 00:04:11.040
execute against in the models.
83
00:04:11.600 --> 00:04:14.300
Oh, they've been cheeky with that one. I've noticed that myself.
84
00:04:14.420 --> 00:04:18.260
I use Perplexity for a lot of our workflows, and I've
85
00:04:18.300 --> 00:04:22.200
noticed that it's extending my conversation on purpose
86
00:04:23.400 --> 00:04:27.280
to add more steps. And I can see, obviously, the tokenization
87
00:04:27.320 --> 00:04:31.250
in the usage. We've got skills that we've
88
00:04:31.260 --> 00:04:34.980
built, organizational skills in the back end, and I actually optimize every single
89
00:04:35.020 --> 00:04:37.460
one to, say, use as little credits as possible.
90
00:04:38.700 --> 00:04:42.520
And it did, and it does. And it was really interesting to see the cost
91
00:04:42.540 --> 00:04:46.300
savings immediately just by me saying, "Find the
92
00:04:46.340 --> 00:04:50.130
most efficient way to run all these skills and automations and
93
00:04:50.200 --> 00:04:54.100
use the least credits as possible." And I was like, "Oh, it actually did
94
00:04:54.140 --> 00:04:54.220
it,"
95
00:04:55.360 --> 00:04:56.140
which was super fascinating.
96
00:04:56.160 --> 00:04:59.820
And just when you did that, they came out, and they decided, "Well, you know what?
97
00:04:59.900 --> 00:05:03.520
We're going to double the cost of your tokens now that you're a captive user" So
98
00:05:04.140 --> 00:05:06.039
I'm not sure you could really ever win that race.
99
00:05:06.060 --> 00:05:06.320
No.
100
00:05:07.140 --> 00:05:07.240
Yeah.
101
00:05:07.280 --> 00:05:10.440
But perception-wise, you might feel pretty good about yourself for a period of
102
00:05:10.520 --> 00:05:10.820
time.
103
00:05:11.460 --> 00:05:11.960
It's true.
104
00:05:13.020 --> 00:05:14.440
Before we go any further, tell everyone
105
00:05:15.280 --> 00:05:18.600
a little bit about your background and your current role, and then we'll kind of
106
00:05:18.640 --> 00:05:18.980
jump in.
107
00:05:20.580 --> 00:05:24.370
Sure. I've been very fortunate to be part of Dayforce for over five
108
00:05:24.400 --> 00:05:24.850
years now,
109
00:05:25.980 --> 00:05:29.590
leading really the push into deep people analytics,
110
00:05:30.120 --> 00:05:30.960
data, AI,
111
00:05:32.300 --> 00:05:35.780
for that period of time. So my two key roles are
112
00:05:35.840 --> 00:05:36.700
basically
113
00:05:37.760 --> 00:05:40.860
the senior vice president of platform engineering.
114
00:05:40.930 --> 00:05:44.760
I look at it this way. I build the aircraft carriers that
115
00:05:44.820 --> 00:05:48.720
planes take off from and land on to make it really a lot easier
116
00:05:48.760 --> 00:05:52.600
for our application teams to be able to build what they need in the
117
00:05:52.620 --> 00:05:56.580
Dayforce suite. And then my other critical role is chief AI
118
00:05:56.640 --> 00:06:00.120
officer. So that includes all the AI governance across our
119
00:06:00.220 --> 00:06:04.160
product, working with our external counsels, working very closely
120
00:06:04.200 --> 00:06:05.630
with our chief privacy officer,
121
00:06:06.460 --> 00:06:09.980
really to make sure that we're answering the questions of can we versus should we.
122
00:06:10.060 --> 00:06:14.038
So really looking at that whole governance structure And a lot of
123
00:06:14.088 --> 00:06:17.728
the AI capabilities that are built today are built within my team,
124
00:06:17.908 --> 00:06:20.548
so we have a great overview of that.
125
00:06:21.008 --> 00:06:23.828
But we also support the corporate side as well, Chris.
126
00:06:24.108 --> 00:06:27.528
So a lot of the purchases that are being made on the corporate side are actually
127
00:06:27.548 --> 00:06:31.368
put through our governance process because of the expertise on my team,
128
00:06:31.828 --> 00:06:35.788
to actually make sure that we're buying software from vendors that
129
00:06:35.888 --> 00:06:39.088
actually are following the right rules and processes
130
00:06:39.948 --> 00:06:43.058
on top of that. But, yeah, simplistically, my
131
00:06:43.608 --> 00:06:46.228
career, including AI, goes back about 20-plus years.
132
00:06:46.668 --> 00:06:46.908
Yeah.
133
00:06:46.988 --> 00:06:50.628
So this isn't new. We've been feeding this
134
00:06:50.688 --> 00:06:54.548
animal for a long time right now. What's really changed is
135
00:06:54.568 --> 00:06:58.468
the whole advent of, obviously, generative AI a
136
00:06:58.508 --> 00:07:02.268
few years ago, and then the speed in which the technology has continued to
137
00:07:02.308 --> 00:07:06.228
accelerate. So it's been a heck of a ride so far, and
138
00:07:06.868 --> 00:07:09.168
it continues to be one as we go forward.
139
00:07:09.228 --> 00:07:10.168
Yeah, it's exciting times.
140
00:07:11.088 --> 00:07:13.648
From your perspective, how is AI
141
00:07:13.768 --> 00:07:17.728
changing the HR function today? Every HR leader
142
00:07:17.788 --> 00:07:21.748
I'm interviewing and speaking to are talking about the new AI HR
143
00:07:21.828 --> 00:07:25.208
operating model of the future and how this is changing and evolving.
144
00:07:25.238 --> 00:07:28.548
I'd love to hear your thoughts and perspective.
145
00:07:28.588 --> 00:07:30.568
I think it's many different things.
146
00:07:30.608 --> 00:07:34.528
I think on the people side alone, it's a massive change management issue
147
00:07:34.568 --> 00:07:38.198
for HR. So outside of the technology, what AI is causing in
148
00:07:38.268 --> 00:07:41.418
that changing in job roles and things of that nature is
149
00:07:41.728 --> 00:07:45.688
really this requirement for HR to be
150
00:07:45.748 --> 00:07:48.828
accelerating change management, but there to support it.
151
00:07:48.948 --> 00:07:52.428
And I think that's probably one of the most critical roles the
152
00:07:52.468 --> 00:07:55.688
CHRO and HR organizations have today, is actually
153
00:07:55.728 --> 00:07:59.638
guiding their organizations through it, because a lot of them don't have a
154
00:07:59.668 --> 00:08:03.518
plan. They're still working and rationalizing,
155
00:08:04.408 --> 00:08:07.868
what's the right group to get together? What's our governance structure?
156
00:08:08.268 --> 00:08:12.128
How do we make it safe for employees to actually not have a fear of missing out or
157
00:08:12.188 --> 00:08:14.688
a fear of messing up, and all of these things.
158
00:08:14.828 --> 00:08:18.668
And on top of that, the angst that I think a lot of the people are feeling when we
159
00:08:18.708 --> 00:08:22.008
look at the potential of how AI can affect
160
00:08:22.648 --> 00:08:26.528
probably 90% of all jobs on the global basis will, in some way,
161
00:08:26.568 --> 00:08:28.168
shape, or form, be influenced by that.
162
00:08:28.628 --> 00:08:28.768
Yeah.
163
00:08:29.148 --> 00:08:32.889
So I think that's a big part of the non-technical component of it.
164
00:08:33.578 --> 00:08:36.398
Then I think there's actually how AI is
165
00:08:36.468 --> 00:08:39.708
impacting their business, the business that they're
166
00:08:39.728 --> 00:08:43.108
delivering to their constituents within the organization.
167
00:08:43.768 --> 00:08:46.148
And I think that's another one that
168
00:08:47.148 --> 00:08:51.088
when I look at it, there's a big difference between sprinkling some AI on a problem
169
00:08:51.148 --> 00:08:53.828
to get 5% benefit and then completely rethinking it.
170
00:08:54.338 --> 00:08:58.118
And I think for a lot of those leaders, they're in that process of really
171
00:08:58.128 --> 00:09:01.968
rationalizing, okay, do we sprinkle it or do we really step back and
172
00:09:01.988 --> 00:09:05.488
figure out how we can really rethink how we deliver our business-
173
00:09:05.917 --> 00:09:05.998
Mm-hmm
174
00:09:06.508 --> 00:09:07.968
... with AI at its core?
175
00:09:08.388 --> 00:09:08.507
Yeah.
176
00:09:08.688 --> 00:09:11.528
And those are some big choices for those leaders.
177
00:09:11.948 --> 00:09:15.618
Yeah. I think like many transformations, it always feels like
178
00:09:15.648 --> 00:09:18.108
HR is brought in too late.
179
00:09:20.137 --> 00:09:20.137
Right.
180
00:09:20.137 --> 00:09:20.137
Yeah.
181
00:09:20.188 --> 00:09:20.488
And then we-
182
00:09:20.568 --> 00:09:22.948
Yeah, I think they have the most important role at the table.
183
00:09:22.958 --> 00:09:25.948
I truly believe that, because of those dynamics.
184
00:09:26.448 --> 00:09:30.108
What do you think the impact is if organizations don't bring
185
00:09:30.288 --> 00:09:32.348
HR in on that journey?
186
00:09:33.928 --> 00:09:35.888
Well, I think there's a number of different ones.
187
00:09:36.408 --> 00:09:39.508
The one that strikes me initially is that
188
00:09:40.308 --> 00:09:44.068
the organizations looking to HR, when you think of career
189
00:09:44.088 --> 00:09:45.428
ladders, career growth,
190
00:09:46.248 --> 00:09:49.288
what does a career look like? How do roles change?
191
00:09:49.948 --> 00:09:53.508
I think so many employees are going to be looking to the HR organization,
192
00:09:53.888 --> 00:09:57.528
not to solve all of those for them, but to really help them understand the
193
00:09:57.588 --> 00:09:59.108
path that they're supposed to be on.
194
00:09:59.568 --> 00:10:02.768
And I think for that reason, it's really a difficult problem
195
00:10:03.928 --> 00:10:07.878
for them to actually tackle. And so for me, when I
196
00:10:07.948 --> 00:10:11.928
see that, that's almost job one. The more core
197
00:10:11.988 --> 00:10:15.588
thing I think that really is important from their perspective is
198
00:10:16.008 --> 00:10:17.588
I actually see them leading the charge.
199
00:10:17.668 --> 00:10:21.148
So getting that chief privacy officer together with, let's say, the
200
00:10:21.228 --> 00:10:25.128
CIO, and coming up with a safe environment in which
201
00:10:25.168 --> 00:10:29.048
the employees can even begin experimenting and be really more
202
00:10:29.128 --> 00:10:33.107
confident in what they're doing. I think we're long past the point
203
00:10:33.128 --> 00:10:36.448
of testing this. This has definitely got real
204
00:10:36.528 --> 00:10:40.068
applications, and I think it requires a very disciplined
205
00:10:40.128 --> 00:10:41.288
approach for that reason.
206
00:10:41.628 --> 00:10:45.528
Yeah. What does that relationship look like with your own
207
00:10:45.548 --> 00:10:49.468
CHRO and yourself and, for perhaps maybe for people
208
00:10:49.488 --> 00:10:52.348
listening that maybe don't have a stronger connection and bond like yourselves,
209
00:10:52.408 --> 00:10:54.128
what advice would you give to them?
210
00:10:55.708 --> 00:10:59.508
I think we've been very fortunate with Amy as our CHRO leading the
211
00:10:59.548 --> 00:10:59.928
charge.
212
00:11:00.868 --> 00:11:03.188
Good or bad, and I think it's a great thing,
213
00:11:04.288 --> 00:11:06.708
they are customer zero for our product.
214
00:11:06.808 --> 00:11:10.608
So in some ways, they're living on the edge of what we're building AI-wise,
215
00:11:11.228 --> 00:11:13.108
and they're very progressive in their thinking.
216
00:11:13.268 --> 00:11:17.108
They're very vocal, which I think is a key part of it as we work
217
00:11:17.148 --> 00:11:20.728
through it. So I think that's where Amy and her team really help
218
00:11:20.738 --> 00:11:24.428
drive us the right way as it relates to that kind of approach.
219
00:11:25.728 --> 00:11:29.388
I think in the broader context, there's an operating group that includes
220
00:11:29.408 --> 00:11:31.408
myself, Amy, our chief
221
00:11:32.488 --> 00:11:36.168
digital officer, Carrie Rasmussen, as well as our chief
222
00:11:36.208 --> 00:11:39.868
privacy officer and CISO. And because of our
223
00:11:39.988 --> 00:11:43.728
size, we're just under 10,000 people, we meet on a
224
00:11:43.748 --> 00:11:47.268
regular basis to make sure internal training, what's the
225
00:11:47.308 --> 00:11:50.728
velocity of that? How are we actually supporting our
226
00:11:50.748 --> 00:11:54.508
employees with more ad hoc but regular meeting groups to
227
00:11:54.568 --> 00:11:56.788
talk about what's succeeding, what's not succeeding?
228
00:11:57.508 --> 00:12:01.208
And I think a real important part of that, selfishly, on the product side, is
229
00:12:01.728 --> 00:12:05.588
how we go through this kind of circle of life from ideation all the
230
00:12:05.628 --> 00:12:07.888
way to delivery of product capabilities,
231
00:12:08.768 --> 00:12:12.468
and watch how our own organization is absorbing them, where they're succeeding,
232
00:12:12.588 --> 00:12:14.648
where maybe they're not, and they need to be improved.
233
00:12:15.168 --> 00:12:18.668
So I think it really provides us with a great lens
234
00:12:18.748 --> 00:12:20.828
into our customers
235
00:12:21.768 --> 00:12:25.516
from that point of view. So I think- All those things combined have been
236
00:12:25.816 --> 00:12:29.616
really, really critical. The sandboxes, though, that we
237
00:12:29.676 --> 00:12:33.406
can all play in successfully, whether we're doing agent development of our own and
238
00:12:33.476 --> 00:12:37.276
things like that, those are critical pieces that our
239
00:12:37.416 --> 00:12:38.576
HR group and our
240
00:12:39.496 --> 00:12:43.456
digital officer have introduced that I think actually give people a lot of
241
00:12:43.516 --> 00:12:47.296
confidence. And that's what's really needed right now, which is a healthy
242
00:12:47.336 --> 00:12:50.136
dose of confidence in learning this new capability.
243
00:12:50.436 --> 00:12:54.336
Yeah. Would you say, in some ways, that AI's brought you closer
244
00:12:54.376 --> 00:12:56.106
together than you were before?
245
00:12:58.316 --> 00:12:59.516
That's interesting. I have-
246
00:12:59.756 --> 00:12:59.845
Okay
247
00:12:59.916 --> 00:13:03.096
... AI that sounds like me, if you want to talk to me or get an article
248
00:13:03.136 --> 00:13:04.906
potentially- ... pursued in how I'd write it.
249
00:13:05.416 --> 00:13:07.025
I'm not sure that brings us closer together,
250
00:13:08.396 --> 00:13:12.345
but it's interesting. I think it's an interesting point, Chris, because
251
00:13:12.456 --> 00:13:12.746
I think it
252
00:13:13.676 --> 00:13:14.956
is a common
253
00:13:18.216 --> 00:13:22.056
need that we all have, and I think it's an interesting
254
00:13:22.076 --> 00:13:25.536
thing to watch psychologically. I see some people that come
255
00:13:25.576 --> 00:13:29.456
together as a group and embrace it because we're all muddling through, let's say,
256
00:13:29.556 --> 00:13:32.016
at the same time, trying to find our way.
257
00:13:32.456 --> 00:13:36.436
But I also see another side of that, which is if I'm defined by
258
00:13:36.756 --> 00:13:39.416
how effectively I use AI at my job,
259
00:13:40.756 --> 00:13:44.656
I actually may want to go off to the side and do more of my own work than share.
260
00:13:44.666 --> 00:13:47.316
So I think both things can be actually true,
261
00:13:48.216 --> 00:13:49.716
and I think that's why we've even seen the
262
00:13:51.516 --> 00:13:55.316
increase in shadow AI and things of that nature in organizations, because
263
00:13:55.856 --> 00:13:59.276
I think people feel somewhat protective about maybe what they're learning or
264
00:13:59.336 --> 00:14:03.076
discovering, which is probably hindering them in the long run, but in the
265
00:14:03.136 --> 00:14:07.036
short run might seem like the right thing to protect what it is that
266
00:14:07.096 --> 00:14:07.575
they do.
267
00:14:08.196 --> 00:14:11.586
Yeah. It's so interesting. What was the word you used there? Shadow... What was it?
268
00:14:11.636 --> 00:14:12.356
Shadow AI.
269
00:14:12.536 --> 00:14:16.416
Shadow AI, yeah. So is that the idea of people using their own
270
00:14:16.496 --> 00:14:20.336
AI tools and ways of working without the business knowing,
271
00:14:20.756 --> 00:14:21.196
essentially?
272
00:14:22.196 --> 00:14:25.936
What a great definition. So yes. How I basically view it
273
00:14:25.996 --> 00:14:29.656
is we're asking people to do a lot, but if you're not providing the
274
00:14:29.716 --> 00:14:33.396
sandboxes, if you're not providing the guardrails, if you don't have something as
275
00:14:33.456 --> 00:14:36.116
simple that says, "This kind of content can go
276
00:14:36.416 --> 00:14:39.276
into this environment and this can't,"
277
00:14:41.556 --> 00:14:45.236
but yet you're insisting on people using AI to actually improve the
278
00:14:45.256 --> 00:14:48.716
effectiveness of what they do, then you're almost pushing it to the edge
279
00:14:48.836 --> 00:14:51.336
there or moving it into the shadows, to your point.
280
00:14:51.896 --> 00:14:54.876
And, if I have a mobile device, I can do it.
281
00:14:55.156 --> 00:14:57.516
If I have my own work laptop, I can do it.
282
00:14:57.616 --> 00:15:01.566
So I would say, actually, those individuals are
283
00:15:01.616 --> 00:15:03.896
really interesting. Don't push them away.
284
00:15:04.016 --> 00:15:07.586
Bring them into you, because they're probably doing some very interesting things
285
00:15:07.616 --> 00:15:11.126
with AI that although there may be better ways to do it that
286
00:15:11.176 --> 00:15:15.116
are important for the business and the business's customers, privacy and
287
00:15:15.176 --> 00:15:18.756
things related to that, at the same time, they're probably going through some
288
00:15:18.796 --> 00:15:21.216
interesting learning that could benefit the organization.
289
00:15:21.296 --> 00:15:25.256
So try to embrace those individuals and bring them back in and give them the safe
290
00:15:25.396 --> 00:15:29.356
space internally to do that versus feeling like they have to go externally.
291
00:15:29.456 --> 00:15:32.916
I think it's probably one of the key things organizations can be doing today.
292
00:15:33.276 --> 00:15:36.576
Yeah. You nailed it there about we have to create that
293
00:15:36.736 --> 00:15:40.556
psychological safety, right? The psychological safety piece that you
294
00:15:40.876 --> 00:15:43.056
mentioned is really important, right?
295
00:15:43.236 --> 00:15:47.056
Because even in my small business, I've found in pockets different people
296
00:15:47.276 --> 00:15:50.696
in our team members doing some really creative things-
297
00:15:51.376 --> 00:15:51.396
Mm
298
00:15:51.496 --> 00:15:54.776
... with AI, right? But sometimes that can be
299
00:15:55.536 --> 00:15:59.076
seen as, oh, they're being lazy because they're using AI.
300
00:15:59.196 --> 00:16:03.176
But I'm like, "Wow, they found a way to, something that was taking them 10 hours,
301
00:16:03.236 --> 00:16:04.996
they've now cut it down to two. That's not lazy.
302
00:16:05.076 --> 00:16:08.476
That's great, and I would want you to share that with the rest of the team." But
303
00:16:08.496 --> 00:16:08.816
unless you-
304
00:16:09.116 --> 00:16:11.336
But what's their incentive to share that, though, Chris?
305
00:16:11.476 --> 00:16:12.196
Yeah, especially if they
306
00:16:13.676 --> 00:16:17.516
are incentivized, monetarily, whatever it may
307
00:16:17.556 --> 00:16:19.936
be, to not do that, right? So it's also like-
308
00:16:19.956 --> 00:16:19.966
Mm
309
00:16:19.996 --> 00:16:22.086
... and that goes back to your point earlier, right?
310
00:16:22.876 --> 00:16:26.326
Last year, I saw a lot of companies to, I think the way you described it is
311
00:16:26.516 --> 00:16:30.366
sprinkle on AI, right? But they didn't think about, well,
312
00:16:30.436 --> 00:16:34.306
how does the work change itself? Let's fundamentally take a step back
313
00:16:34.336 --> 00:16:38.256
and look at not just adding AI on top of existing processes,
314
00:16:38.276 --> 00:16:42.076
ways of working, but taking the opportunity to fundamentally rethink,
315
00:16:42.496 --> 00:16:42.696
right?
316
00:16:42.816 --> 00:16:43.036
Mm-hmm.
317
00:16:43.316 --> 00:16:46.836
Because if you're adding AI onto something, but you're incentivizing someone to go
318
00:16:46.876 --> 00:16:47.726
in this direction,
319
00:16:49.096 --> 00:16:52.276
but your AI's doing X, you're going to
320
00:16:52.956 --> 00:16:56.456
just... I feel like now I'm looking at more like how do I take away,
321
00:16:56.496 --> 00:16:58.156
remove friction, like-
322
00:16:58.256 --> 00:16:58.266
Mm-hmm
323
00:16:58.266 --> 00:17:00.516
... before I start thinking about technology.
324
00:17:00.876 --> 00:17:02.716
Technology comes secondary.
325
00:17:03.276 --> 00:17:03.336
Yeah.
326
00:17:03.436 --> 00:17:05.206
I don't know if that makes sense. But yeah.
327
00:17:05.206 --> 00:17:08.206
It does. I think the only thing that I would add to that is
328
00:17:08.336 --> 00:17:11.477
that governance, especially for what we do,
329
00:17:11.737 --> 00:17:14.156
selfishly, has to be at the forefront.
330
00:17:15.856 --> 00:17:19.536
I may want to completely rebuild how I think payroll or workforce
331
00:17:19.616 --> 00:17:23.596
management or engagement might work, but arguably speaking, there's
332
00:17:23.696 --> 00:17:25.306
a whole series of-
333
00:17:25.556 --> 00:17:25.615
Sure
334
00:17:25.656 --> 00:17:29.186
... rules and regulations on a global basis that don't care about what you think is
335
00:17:29.236 --> 00:17:31.116
the neatest new way to consider it.
336
00:17:31.206 --> 00:17:35.016
So I think it's also understanding that there's that layer,
337
00:17:35.516 --> 00:17:36.696
especially in compliance-
338
00:17:37.096 --> 00:17:37.216
Sure
339
00:17:37.226 --> 00:17:40.056
... that is so critical to an organization's success.
340
00:17:40.476 --> 00:17:44.216
The last thing you want to do is actually rebuild a process that leads you
341
00:17:44.296 --> 00:17:48.036
to actually be fined for not paying people on
342
00:17:48.096 --> 00:17:49.716
time or doing things of that nature.
343
00:17:50.136 --> 00:17:53.476
I don't think that's the intended consequence, but arguably that could be the
344
00:17:53.536 --> 00:17:57.496
unintended consequence of simply rebuilding blindly
345
00:17:57.556 --> 00:17:59.996
a process. So I think it has to be thoughtful, to your point.
346
00:18:00.216 --> 00:18:00.636
Mm-hmm.
347
00:18:01.116 --> 00:18:05.016
But I think there is an opportunity to really rethink
348
00:18:05.316 --> 00:18:06.276
the nature of work
349
00:18:07.176 --> 00:18:11.096
that we're doing and the journeys that we're trying to accomplish using AI
350
00:18:11.256 --> 00:18:13.176
at its core versus just on top.
351
00:18:13.756 --> 00:18:17.696
Yeah. I think one of the things that we're hearing in a lot of the conversations
352
00:18:17.736 --> 00:18:21.636
in our CHRO roundtables we do on a weekly basis in person is, how can
353
00:18:21.676 --> 00:18:24.696
HR leaders balance AI-driven efficiency
354
00:18:25.556 --> 00:18:26.856
with the need for the human
355
00:18:27.676 --> 00:18:31.076
oversight and the human in the loop at the same time, right?
356
00:18:33.512 --> 00:18:37.472
The whole concept of whether the human is in, on, or out of the loop is
357
00:18:37.512 --> 00:18:39.972
an important one to consider in any of those processes.
358
00:18:40.712 --> 00:18:44.352
When we look at it from our standpoint, we're very, I
359
00:18:44.432 --> 00:18:48.092
think, thoughtful about it. Our view is that the customer should
360
00:18:48.132 --> 00:18:51.652
decide where they want to be in that process.
361
00:18:51.732 --> 00:18:55.552
We want to enable them to be as heavy touch or light
362
00:18:55.612 --> 00:18:59.492
touch as they choose to be. But arguably, in the end, that's a
363
00:18:59.552 --> 00:19:03.272
customer decision, not specifically our decision on that.
364
00:19:03.302 --> 00:19:07.132
So we enable it through a lot of the tools that we're building to either
365
00:19:07.172 --> 00:19:11.072
be high touch or low touch, depending on where that customer wants
366
00:19:11.112 --> 00:19:11.492
to live,
367
00:19:12.352 --> 00:19:15.992
on that spectrum of, "I'm absolutely going to change the world with AI,"
368
00:19:16.092 --> 00:19:19.922
or, "We're going to go into this gradually, and we're going to understand it and
369
00:19:19.952 --> 00:19:23.572
feel comfortable as we go into it," knowing that more or less, the
370
00:19:23.692 --> 00:19:27.492
capabilities they need to go from one end to the other are there
371
00:19:27.592 --> 00:19:30.232
for them. And I think that relaxes people
372
00:19:31.172 --> 00:19:34.412
to allow them to work at their own pace, because so often,
373
00:19:35.272 --> 00:19:38.572
even today, how many times have you seen a piece of software you're using and all
374
00:19:38.611 --> 00:19:41.612
of a sudden it's like, "Oh, what's that new AI functionality in there?
375
00:19:41.972 --> 00:19:43.752
I didn't know that existed before."
376
00:19:43.792 --> 00:19:43.912
Yeah.
377
00:19:43.952 --> 00:19:46.072
That concept of surprising a customer
378
00:19:47.212 --> 00:19:51.092
is terrible. We believe AI by choice should be, especially again
379
00:19:51.132 --> 00:19:55.062
in the whole compliance area, should be leading the charge so that
380
00:19:55.152 --> 00:19:58.712
I decide if I'm going to use your AI capability.
381
00:19:58.722 --> 00:20:02.512
You don't get to decide for me. And I think that gives confidence to those
382
00:20:02.532 --> 00:20:06.452
individuals then to say, "If I can make that choice, then I can also
383
00:20:06.532 --> 00:20:10.412
decide where we're going to play in, on, or out of the
384
00:20:10.492 --> 00:20:12.112
loop as it relates to that."
385
00:20:12.512 --> 00:20:12.692
Yeah.
386
00:20:12.722 --> 00:20:16.672
And I think it's an important way to really frame thinking about how you're going
387
00:20:16.732 --> 00:20:18.752
to apply AI in your organization.
388
00:20:18.872 --> 00:20:20.112
Yeah. It's really, really important.
389
00:20:20.192 --> 00:20:23.782
On that point, what are some of the frequently asked questions or
390
00:20:23.852 --> 00:20:27.492
conversations that you're having with Dayforce customers, prospects?
391
00:20:27.992 --> 00:20:31.012
Kind of what's the main themes or questions you're hearing right now?
392
00:20:31.792 --> 00:20:35.412
And it's interesting, I just got back from our major Dallas summit,
393
00:20:35.612 --> 00:20:39.282
as well as a very interesting opportunity in New
394
00:20:39.332 --> 00:20:43.192
York City, where I sat down with probably about 40 HR
395
00:20:43.252 --> 00:20:43.571
leaders.
396
00:20:44.972 --> 00:20:48.522
First of all, I think I'm always struck by where literacy is
397
00:20:48.572 --> 00:20:49.282
currently today,
398
00:20:50.372 --> 00:20:53.752
and where most organizations are.
399
00:20:53.812 --> 00:20:57.242
If we read what's in the press, everyone is at the one end of this
400
00:20:57.572 --> 00:21:00.322
spectrum. They're doing everything with AI, and
401
00:21:00.672 --> 00:21:03.332
they're just having a fantastic time.
402
00:21:03.372 --> 00:21:05.772
But when you have a grounded discussion with most of them,
403
00:21:07.212 --> 00:21:11.172
it's probably about a third are actually really moving
404
00:21:11.332 --> 00:21:14.772
on a fairly quick continuum. And the rest
405
00:21:15.432 --> 00:21:18.852
are still trying to figure out even basic governance and things of that nature.
406
00:21:18.932 --> 00:21:22.861
So I think, again, since we're moving so quickly, we assume by the press and
407
00:21:22.892 --> 00:21:25.852
what we see, everything is just flying out the door.
408
00:21:26.132 --> 00:21:29.972
And when the reality is, I think everybody is somewhere on that
409
00:21:30.012 --> 00:21:33.552
line, and a lot of them are just in the, I'd say, the
410
00:21:33.612 --> 00:21:34.652
earlier stages.
411
00:21:35.192 --> 00:21:35.242
Mm-hmm.
412
00:21:35.792 --> 00:21:37.312
And I define that even by literacy.
413
00:21:37.352 --> 00:21:41.312
If you think about it, literacy is probably running about 37, 38%.
414
00:21:42.832 --> 00:21:45.892
So that kind of tells you how far you're actually going to be able to run
415
00:21:46.472 --> 00:21:50.152
when the literacy rates still aren't where they need to be for everybody
416
00:21:50.192 --> 00:21:53.992
to contribute the way they probably would like to in
417
00:21:54.052 --> 00:21:56.092
using the different applications
418
00:21:56.912 --> 00:21:58.992
of AI within their organizations.
419
00:21:59.012 --> 00:21:59.112
Yeah.
420
00:21:59.192 --> 00:22:01.352
So I think that's one of the most important things.
421
00:22:01.772 --> 00:22:04.372
One of the other things that we're starting to see really importantly in these
422
00:22:04.432 --> 00:22:07.312
organizations is the trust factor.
423
00:22:09.572 --> 00:22:13.232
So many organizations are grappling why they're moving slower, is they're grappling
424
00:22:13.292 --> 00:22:17.232
just with whether or not their data quality is strong enough to support
425
00:22:17.292 --> 00:22:20.812
AI. Garbage data, garbage AI, not exactly
426
00:22:20.892 --> 00:22:22.172
a magic
427
00:22:23.112 --> 00:22:25.992
mix of- ... positivity. So
428
00:22:26.952 --> 00:22:30.692
just making sure that that singular platform,
429
00:22:31.032 --> 00:22:34.522
that single view of that employee is
430
00:22:34.632 --> 00:22:38.492
accurate and that we can build AI on top of it, I think is so critical for
431
00:22:38.572 --> 00:22:42.432
many organizations. So it takes them through a process where
432
00:22:42.452 --> 00:22:44.792
they really have to rationalize some of these things
433
00:22:45.512 --> 00:22:47.632
as they move forward.
434
00:22:48.072 --> 00:22:51.072
Yeah. What are you seeing as like the common mistakes that
435
00:22:51.932 --> 00:22:55.131
those organizations are making on that journey, the sort of common pitfalls?
436
00:22:57.932 --> 00:23:01.552
I think some of the first things that I see when I think about just the process
437
00:23:01.612 --> 00:23:05.322
that a lot of them go through is they look at the, there are many opportunities.
438
00:23:05.392 --> 00:23:07.072
We actually have a very well-defined
439
00:23:07.972 --> 00:23:11.642
way of bringing in ideas from across 9,500-plus employees
440
00:23:12.332 --> 00:23:16.311
and actually saying to ourselves, "Okay, first thing is, do we have the
441
00:23:16.352 --> 00:23:20.332
right data to even address the AI opportunity that's here?" And in
442
00:23:20.352 --> 00:23:24.312
many cases, you don't. You may have to go get it somewhere else, but you
443
00:23:24.352 --> 00:23:26.112
may not have it natively in your business.
444
00:23:26.192 --> 00:23:30.112
So that's kind of an on or an off ramp to whether or not you're going to go forward
445
00:23:30.152 --> 00:23:33.972
or not. And then I think another critical question that everyone needs to
446
00:23:34.092 --> 00:23:37.992
ask is, "Okay, so we know that we
447
00:23:38.072 --> 00:23:41.192
actually have the data, but is there a regulatory or
448
00:23:41.952 --> 00:23:45.432
an environment in which we can't use that data that way?" Thinking
449
00:23:45.512 --> 00:23:49.132
of how it introduces higher risk for biasness and things of that
450
00:23:49.152 --> 00:23:52.812
nature. And then, once you pass that gate, then the
451
00:23:52.852 --> 00:23:56.612
question is, well, should we? So there's a should we and can we
452
00:23:56.692 --> 00:24:00.632
conversation about, we could do this, but what's the actual
453
00:24:00.692 --> 00:24:04.572
impact on an individual or organization in
454
00:24:04.612 --> 00:24:07.632
doing so? Those are three really simple questions.
455
00:24:07.672 --> 00:24:11.412
They can be somewhat complex, but that really help an organization frame,
456
00:24:11.972 --> 00:24:12.252
okay,
457
00:24:13.332 --> 00:24:16.902
we're where we need to be. Those are all on-ramps to what we're doing.
458
00:24:16.932 --> 00:24:20.212
And then I think the biggest mistake I see today, too, is
459
00:24:20.612 --> 00:24:23.372
organizations say, "Great, let's run with this." So we have a pilot.
460
00:24:23.892 --> 00:24:27.832
There's lots of pilots going on. But everyone's measuring them like three months
461
00:24:27.932 --> 00:24:29.072
out, six months out.
462
00:24:30.012 --> 00:24:33.571
Well, that's a awfully late period to actually understand if it's working.
463
00:24:33.692 --> 00:24:37.572
So I think the biggest recommendation I've had in working with some of these groups
464
00:24:37.672 --> 00:24:38.012
is-
465
00:24:39.524 --> 00:24:43.384
Every week you should be evaluating whether it works or not, and it's a fast
466
00:24:43.464 --> 00:24:44.984
off-ramp if it's not.
467
00:24:44.994 --> 00:24:44.994
Wow.
468
00:24:45.004 --> 00:24:48.484
And I think that's how we've looked at it, even in our product itself, which is
469
00:24:49.384 --> 00:24:53.354
if you're not driving quantifiable value or real value out of the process,
470
00:24:53.904 --> 00:24:54.364
stop it.
471
00:24:55.304 --> 00:24:57.604
It's not going to get necessarily better over time.
472
00:24:57.644 --> 00:25:01.394
So be harsh in stopping those things that aren't working.
473
00:25:01.404 --> 00:25:02.714
And then those that are working,
474
00:25:04.084 --> 00:25:07.524
what you really need to understand is how do you scale that?
475
00:25:07.584 --> 00:25:10.634
It's one thing when you have an individual who's doing something off the side of
476
00:25:10.664 --> 00:25:12.724
their desk, to your point, that works well for them.
477
00:25:13.824 --> 00:25:14.054
But
478
00:25:15.144 --> 00:25:18.984
let's say it works really well. How do you scale that out to 4,000 or 5,000 people?
479
00:25:19.034 --> 00:25:19.554
Mm-hmm.
480
00:25:19.624 --> 00:25:23.464
Or hundreds? And I don't think we give a lot of thought to that
481
00:25:23.544 --> 00:25:27.284
as well. So we have infrastructures and everything else that aren't really meant to
482
00:25:27.364 --> 00:25:29.804
scale, so we're kind of holding ourselves back.
483
00:25:29.904 --> 00:25:33.404
So I think there's some very basic lessons learned that aren't just about
484
00:25:33.764 --> 00:25:36.924
necessarily AI. I think they've existed in all the different
485
00:25:37.984 --> 00:25:41.264
pieces of technology we've seen come in that are more transformational in an
486
00:25:41.324 --> 00:25:44.504
organization. So I think similar rules apply.
487
00:25:45.744 --> 00:25:48.184
I love there was a lot to take in there, and I love it.
488
00:25:48.224 --> 00:25:52.044
I think HR traditionally has been quite risk
489
00:25:52.104 --> 00:25:55.544
adverse in terms of doing, coming in,
490
00:25:55.624 --> 00:25:58.143
experimenting, piloting.
491
00:25:58.174 --> 00:25:58.174
Mm-hmm.
492
00:25:58.724 --> 00:26:02.184
Like you said, three months actually in the past would be seen as
493
00:26:02.684 --> 00:26:04.524
quite a short period.
494
00:26:05.534 --> 00:26:09.284
And now you're saying, "No, actually, we need to create even
495
00:26:09.384 --> 00:26:12.074
shorter and understand that we're not going to get it right every time, and we need
496
00:26:12.124 --> 00:26:16.054
to innovate fast and make decisions." And that's
497
00:26:16.064 --> 00:26:18.304
traditionally not how HR operated, right?
498
00:26:18.384 --> 00:26:19.054
Mm-hmm.
499
00:26:19.054 --> 00:26:20.184
As well. So that's a big shift.
500
00:26:21.164 --> 00:26:21.604
It is.
501
00:26:22.504 --> 00:26:26.484
And I think for everyone, not just HR, but I think just especially HR, because when
502
00:26:26.504 --> 00:26:28.864
you said three months, I thought you was going to say that's a good time, and you
503
00:26:28.944 --> 00:26:29.874
were like, "No, actually, we need
504
00:26:30.994 --> 00:26:33.404
to even shorter." But I'm with you.
505
00:26:33.524 --> 00:26:37.274
I know even with the tools that we use, within days, quite honestly-
506
00:26:37.874 --> 00:26:38.714
Mm-hmm
507
00:26:38.714 --> 00:26:42.244
... wherever it's actually living up to it, and chatting to the team, and seeing
508
00:26:42.264 --> 00:26:45.284
the usage, and not just the usage, but the output, because usage doesn't tell the
509
00:26:45.364 --> 00:26:45.804
story.
510
00:26:46.544 --> 00:26:47.204
Yep.
511
00:26:47.223 --> 00:26:50.264
You could see amazing usage and no outcomes.
512
00:26:50.744 --> 00:26:52.204
So that doesn't really...
513
00:26:53.084 --> 00:26:56.964
On another side, it's a bit overwhelming for many of the HR leaders I
514
00:26:57.064 --> 00:27:00.804
speak to. So as someone who's sitting in your seat, when they're talking to other
515
00:27:00.904 --> 00:27:04.024
vendors and AI providers or partners, what are some of the
516
00:27:04.544 --> 00:27:08.404
questions that they have to ask, that they should be asking on those
517
00:27:08.444 --> 00:27:11.024
calls? Because everyone claims they've got the silver bullet, and they've got the
518
00:27:11.084 --> 00:27:13.924
best model, and they can solve all your problems and integrate with everything,
519
00:27:14.264 --> 00:27:16.224
right? Everyone tells that story.
520
00:27:17.784 --> 00:27:21.664
Well, I think it depends. We see it on a global basis, so
521
00:27:21.724 --> 00:27:25.564
whether it's you're operating in the UK, the EU,
522
00:27:25.624 --> 00:27:29.044
APJ, or US, Canada, or South America,
523
00:27:29.544 --> 00:27:33.484
all the biggest first challenge is the regulatory environment is a
524
00:27:33.564 --> 00:27:34.584
patchwork right now.
525
00:27:34.764 --> 00:27:34.844
Mm-hmm.
526
00:27:34.884 --> 00:27:38.804
And it's constantly changing. And in the US, for example,
527
00:27:38.934 --> 00:27:42.624
where there's no federal legislation around the use of AI,
528
00:27:42.764 --> 00:27:46.714
that's left it to the states to all figure out, "Okay, what
529
00:27:46.724 --> 00:27:50.244
are we going to do that's good for California or good for Colorado?" So I think
530
00:27:50.304 --> 00:27:53.714
the first biggest challenge is actually understanding the environment that you're
531
00:27:53.764 --> 00:27:57.524
actually working with because the issues I've
532
00:27:57.604 --> 00:28:01.124
heard you talk about before around
533
00:28:01.164 --> 00:28:05.144
biasness in decision-making and things related to that have real impacts.
534
00:28:05.164 --> 00:28:09.064
That's why HR tends to be more self-conscious about those
535
00:28:09.104 --> 00:28:12.724
choices and using those kind of capabilities, because the
536
00:28:12.744 --> 00:28:14.804
potential for adverse impacts is real.
537
00:28:15.204 --> 00:28:15.744
Yes.
538
00:28:15.804 --> 00:28:19.244
And so I think that's what causes that caution for the right
539
00:28:19.304 --> 00:28:23.204
reasons, and why when I'm thinking about what you want to
540
00:28:23.244 --> 00:28:27.064
talk to a vendor about, the first thing is are you
541
00:28:27.084 --> 00:28:27.964
doing the basics?
542
00:28:29.224 --> 00:28:32.624
Is it AI by choice? I know how you're
543
00:28:32.744 --> 00:28:36.484
using particular AI capabilities, and I actually
544
00:28:36.544 --> 00:28:40.524
opt in, not opt out. And I think that's really important
545
00:28:40.604 --> 00:28:44.184
because the second an employee sees an opportunity to use AI in a piece of
546
00:28:44.204 --> 00:28:48.064
software, even if you didn't opt into it, that's almost like
547
00:28:48.124 --> 00:28:52.004
signaling because it's inside my walled garden of my company, it's okay for me to
548
00:28:52.044 --> 00:28:55.983
go use that when it may not be. So I think the first most important thing
549
00:28:56.064 --> 00:28:59.824
is are you able to opt in, not simply have to opt
550
00:28:59.923 --> 00:29:03.584
out? I think the next thing really is what are you doing with my data?
551
00:29:04.704 --> 00:29:07.984
I think on that basis alone, are you training off my data?
552
00:29:08.004 --> 00:29:11.984
How are you using my data? Where is it at rest from a sovereignty
553
00:29:12.064 --> 00:29:12.624
point of view?
554
00:29:12.664 --> 00:29:12.674
Mm-hmm.
555
00:29:12.764 --> 00:29:15.924
All of those things are a very important consideration for
556
00:29:15.964 --> 00:29:19.124
organizations around that particular area.
557
00:29:19.564 --> 00:29:23.084
I think also when I look at it, there's some very basic principles
558
00:29:23.184 --> 00:29:27.024
around consent and things of that nature, that
559
00:29:27.124 --> 00:29:29.654
I'm constantly surprised at, that
560
00:29:30.624 --> 00:29:33.504
many software companies think of as an afterthought.
561
00:29:33.534 --> 00:29:37.284
"Oh yeah, we'll put that on the product roadmap, maybe for later. That's great.
562
00:29:37.324 --> 00:29:40.574
We'll have informed consent." Well, no, you can't because the regulatory
563
00:29:40.604 --> 00:29:42.284
environment in the EU requires it.
564
00:29:42.764 --> 00:29:46.304
So it's actually looking for organizations from that standpoint.
565
00:29:46.704 --> 00:29:50.604
The last one is governance and trust, which is a big part
566
00:29:50.664 --> 00:29:54.424
of it. I know that we've placed a huge focus on our
567
00:29:54.444 --> 00:29:57.724
certifications around the ISO 42001 standard,
568
00:29:58.764 --> 00:30:01.664
on the work we've done for NIST attestation.
569
00:30:01.704 --> 00:30:05.504
For us, we have an external ethics board made up of
570
00:30:06.024 --> 00:30:09.364
not-for-profit as well as very esteemed
571
00:30:09.424 --> 00:30:13.144
individuals in education and things of that nature that we work
572
00:30:13.184 --> 00:30:16.444
with to answer the can we versus should we question.
573
00:30:17.304 --> 00:30:20.764
All of those things as just part of our governance, I think is what
574
00:30:21.644 --> 00:30:25.224
I would look for, and what we do look for when we're helping and assisting our
575
00:30:25.284 --> 00:30:29.134
chief digital officer's team in evaluating software.
576
00:30:29.604 --> 00:30:32.924
All those things are part of it. And sadly,
577
00:30:32.934 --> 00:30:36.174
most vendors we talk to struggle with most of those
578
00:30:36.224 --> 00:30:37.224
questions.
579
00:30:37.284 --> 00:30:37.524
Yeah.
580
00:30:37.944 --> 00:30:41.854
Or frankly, the people that you're talking to simply are clueless, and they don't
581
00:30:41.884 --> 00:30:43.024
know.
582
00:30:43.234 --> 00:30:44.674
Yep. I've talked to a few of those people
583
00:30:46.284 --> 00:30:49.724
over the time, and I'm like, "Wait a minute. I know more than you about this.
584
00:30:49.764 --> 00:30:52.344
This is scary, and you're supposed to be answering my questions."
585
00:30:53.676 --> 00:30:56.696
Yeah. I was going to ask you about the AI compliance certification side, because I
586
00:30:56.716 --> 00:30:57.816
know you've done a lot of work there,
587
00:30:59.276 --> 00:31:01.016
especially around ethical HR technology.
588
00:31:01.036 --> 00:31:01.046
Mm-hmm.
589
00:31:01.046 --> 00:31:01.896
And we've already seen
590
00:31:02.996 --> 00:31:04.256
what happens when you don't do that right.
591
00:31:05.096 --> 00:31:06.096
We're not going to name any names.
592
00:31:06.896 --> 00:31:08.636
No, we've seen lots of examples-
593
00:31:09.076 --> 00:31:09.156
Yeah.
594
00:31:09.166 --> 00:31:10.416
... of how not to do it right.
595
00:31:10.536 --> 00:31:14.116
And we see the implications because, AI is great, but also that
596
00:31:14.136 --> 00:31:16.816
implication will scale just as fast, right?
597
00:31:16.896 --> 00:31:17.396
If you
598
00:31:18.336 --> 00:31:21.956
don't do that, the risks are also exponential
599
00:31:24.036 --> 00:31:27.396
as well. So, it's really interesting. Yeah.
600
00:31:27.596 --> 00:31:30.556
One other thing that came up recently, and I would love to hear, I'm noticing it
601
00:31:30.616 --> 00:31:34.496
more and more. For example, I was speaking with Nicole Lamoureux, CTO for
602
00:31:34.556 --> 00:31:34.976
IBM-
603
00:31:35.716 --> 00:31:35.786
Mm-hmm
604
00:31:35.816 --> 00:31:37.716
... good friend of ours, and
605
00:31:39.016 --> 00:31:42.776
they've spent almost probably years at this point,
606
00:31:43.276 --> 00:31:46.296
completely, because of AI, reevaluating their
607
00:31:46.336 --> 00:31:50.196
entire tech stack. I think they went from over 1,000
608
00:31:50.256 --> 00:31:53.536
systems to less than 100 in a year.
609
00:31:53.606 --> 00:31:53.646
Mm-hmm.
610
00:31:55.156 --> 00:31:56.406
What are your thoughts on that, though?
611
00:31:56.456 --> 00:32:00.236
Because obviously, the current tools and systems are not built
612
00:32:00.376 --> 00:32:03.556
AI native. There's this sort of patchwork happening,
613
00:32:04.616 --> 00:32:08.496
like I call them AI stickers or AI bolt-ons that we're seeing in products.
614
00:32:08.556 --> 00:32:08.686
Mm-hmm.
615
00:32:09.136 --> 00:32:11.846
How are you and the team viewing this, if that makes sense?
616
00:32:12.816 --> 00:32:14.636
I think it's actually a very interesting question.
617
00:32:14.996 --> 00:32:18.866
Looking at it from Dayforce's point of view, one of the big things we've talked
618
00:32:18.936 --> 00:32:22.366
about a lot over the last probably two years
619
00:32:22.936 --> 00:32:26.576
plus, given the length of time passed then that we've actually been involved with
620
00:32:26.636 --> 00:32:30.476
AI, is this concept of 12 to one. There are many
621
00:32:30.516 --> 00:32:34.416
different systems that power an HR organization, and
622
00:32:34.456 --> 00:32:38.136
therefore the overarching organization, whether it's recruiting, workforce
623
00:32:38.276 --> 00:32:40.256
planning, strategic workforce planning,
624
00:32:41.076 --> 00:32:43.636
payroll, engagement, and things of that nature.
625
00:32:44.126 --> 00:32:47.556
And typically, many organizations, those pieces were scattered everywhere.
626
00:32:47.596 --> 00:32:47.776
Mm-hmm.
627
00:32:47.856 --> 00:32:51.816
And so this goes back a little bit to our earlier conversation.
628
00:32:51.896 --> 00:32:55.436
If your data is everywhere, the quality of your data as you try to
629
00:32:55.476 --> 00:32:59.336
transform it and bring it into one place to effectively drive AI on top of
630
00:32:59.416 --> 00:33:03.076
it, becomes very messy, and so the quality of the AI
631
00:33:03.116 --> 00:33:06.936
suffers. So we've really taken this approach of 12 to
632
00:33:07.036 --> 00:33:10.656
one, which really brings all those pieces together with what we
633
00:33:10.676 --> 00:33:14.656
would call that single version of truth, and
634
00:33:14.716 --> 00:33:18.456
that single version of truth is where all your employee data lives in one
635
00:33:18.596 --> 00:33:22.516
place. And our argument, really, from a compliance point of view
636
00:33:22.616 --> 00:33:26.376
is, that compliant layer of data, that pays your
637
00:33:26.396 --> 00:33:29.416
employees, so that's how much you believe in it and how
638
00:33:30.236 --> 00:33:34.156
strong that data layer is, if you have that in place and your systems
639
00:33:34.176 --> 00:33:38.166
are consolidated around that one point of view, then the quality
640
00:33:38.196 --> 00:33:41.976
of the AI that you can then apply across all of those different
641
00:33:42.036 --> 00:33:45.786
systems from almost a journey point of view versus a point solution,
642
00:33:46.296 --> 00:33:47.015
is critical.
643
00:33:47.556 --> 00:33:47.786
Mm-hmm.
644
00:33:47.916 --> 00:33:51.276
That system of outcomes versus just maybe
645
00:33:51.296 --> 00:33:55.156
that one value here, how you use
646
00:33:55.166 --> 00:33:58.436
that to transform the way you look at using your data is really important.
647
00:33:59.076 --> 00:34:02.736
So I think if I look at it, that's the foundational piece.
648
00:34:03.276 --> 00:34:06.176
Even when we work, and we've made a number of
649
00:34:06.256 --> 00:34:10.006
acquisitions on a consistent basis, the first thing we always do
650
00:34:10.776 --> 00:34:14.236
is when we acquire, the first thing we do is integrate that data into that data
651
00:34:14.296 --> 00:34:18.216
model, so there's one place in which to really use that leverage
652
00:34:18.357 --> 00:34:20.417
point to drive better AI.
653
00:34:20.836 --> 00:34:21.116
Yeah.
654
00:34:21.457 --> 00:34:25.136
So having that deliberate approach within an organization
655
00:34:25.336 --> 00:34:29.196
of bringing that data to a place where it's trusted, and then you have that
656
00:34:29.276 --> 00:34:32.957
trust layer over it, because as you well know, Chris, AI doesn't
657
00:34:33.036 --> 00:34:34.026
care what you're feeding it.
658
00:34:34.437 --> 00:34:34.556
No.
659
00:34:34.676 --> 00:34:37.976
It's just going to use it. It doesn't go, "Gee, I wonder if Chris should be using
660
00:34:38.016 --> 00:34:40.216
this data today." I don't think that happens too often.
661
00:34:40.336 --> 00:34:40.476
No.
662
00:34:40.776 --> 00:34:44.417
But because our security layer is over the data before it ever
663
00:34:44.457 --> 00:34:48.196
reaches AI or reporting or analytics, and knows who you
664
00:34:48.296 --> 00:34:52.236
are on a role level, let's say, we always know that
665
00:34:52.397 --> 00:34:56.016
safely, safety's so important, the data going up and through that
666
00:34:56.056 --> 00:34:59.786
layer is always about what you're allowed to see, what you're allowed to work
667
00:34:59.836 --> 00:35:03.416
with in your context. And I think that's another important part that
668
00:35:03.496 --> 00:35:07.336
gives HR organizations real confidence in what
669
00:35:07.356 --> 00:35:07.876
they're doing.
670
00:35:08.236 --> 00:35:11.096
Yeah. No, I think you hit the nail on the head there, and I think that's exactly
671
00:35:11.115 --> 00:35:13.756
what's driving this consolidation right now.
672
00:35:13.836 --> 00:35:14.376
It was already-
673
00:35:14.556 --> 00:35:14.756
Mm-hmm
674
00:35:14.776 --> 00:35:16.136
... a bit of a mess before AI,
675
00:35:17.256 --> 00:35:18.936
let alone with AI, right?
676
00:35:19.676 --> 00:35:19.976
Yes.
677
00:35:20.236 --> 00:35:24.146
And there was always this struggle of how do we have one source of truth and
678
00:35:24.176 --> 00:35:27.976
have our data when you've got our ATS over here, our LMS over there, our
679
00:35:28.136 --> 00:35:29.276
HRS over here-
680
00:35:29.476 --> 00:35:29.506
Mm-hmm
681
00:35:29.506 --> 00:35:32.056
... our learning content systems.
682
00:35:32.576 --> 00:35:36.456
It was just so much, and now every one of those
683
00:35:36.516 --> 00:35:39.076
partners have added their own individual AI.
684
00:35:39.836 --> 00:35:40.476
Yep.
685
00:35:42.416 --> 00:35:45.616
And so now it's creating more chaos
686
00:35:45.916 --> 00:35:47.556
in the system.
687
00:35:48.056 --> 00:35:50.466
Well, it really hurts how you can go across all-
688
00:35:50.476 --> 00:35:50.526
Yeah
689
00:35:50.526 --> 00:35:52.656
... those systems then because it's like
690
00:35:53.556 --> 00:35:57.436
seeing the game Whack-a-Mole, where a mole pops up a hole and you
691
00:35:57.856 --> 00:35:59.556
hit it on the head. Now you've got all of these.
692
00:35:59.576 --> 00:36:03.566
But if you have a consolidated view of it, that's
693
00:36:03.576 --> 00:36:06.886
where, again, the difference between the sprinkling and the transforming the way
694
00:36:06.936 --> 00:36:10.896
you think of your business is going to take off
695
00:36:10.996 --> 00:36:14.716
in that regard. And so it doesn't mean that there's not some very good
696
00:36:14.736 --> 00:36:18.596
solutions out there from that standpoint, but arguably speaking, the
697
00:36:18.656 --> 00:36:21.996
more that you introduce that way, the harder it is to manage from a governance
698
00:36:22.096 --> 00:36:22.516
point of view.
699
00:36:23.096 --> 00:36:26.396
I think more importantly as well, the biggest impact is the impact on employee
700
00:36:26.436 --> 00:36:27.016
experience.
701
00:36:28.256 --> 00:36:28.536
Yeah.
702
00:36:29.456 --> 00:36:30.796
It's a great example because-
703
00:36:30.876 --> 00:36:30.886
Right
704
00:36:30.886 --> 00:36:33.016
... it's disjointed then, in that case. Yeah.
705
00:36:33.076 --> 00:36:33.255
Yeah.
706
00:36:33.436 --> 00:36:34.406
Very good point.
707
00:36:34.576 --> 00:36:36.076
It's overwhelming. It's like-
708
00:36:36.196 --> 00:36:36.206
Mm-hmm
709
00:36:36.206 --> 00:36:38.396
... I've got to go over here, I've got like 20 agents.
710
00:36:40.116 --> 00:36:43.756
It's becoming very difficult for organizations and
711
00:36:44.096 --> 00:36:48.036
employees to know where they're going because every tool has an agent, and every
712
00:36:48.096 --> 00:36:50.816
tool has some sort of a different entry point-
713
00:36:51.656 --> 00:36:51.876
Mm-hmm
714
00:36:52.216 --> 00:36:55.822
... in as well. It's interesting because my next
715
00:36:55.832 --> 00:36:59.592
question was I going to ask you around like what would your
716
00:36:59.612 --> 00:37:02.892
advice be for HR leaders to future-proof their AI strategy?
717
00:37:02.912 --> 00:37:05.992
Which is crazy to ask that question because many of them are just starting
718
00:37:07.632 --> 00:37:11.592
with their AI. So talking about future-proofing maybe is a step ahead, but I
719
00:37:11.612 --> 00:37:14.052
would love to hear your thoughts and perspectives on that.
720
00:37:15.532 --> 00:37:18.252
Ah, future-proofing their AI strategy.
721
00:37:18.372 --> 00:37:21.852
Well, with things changing the way they are, I would definitely say that
722
00:37:22.232 --> 00:37:25.852
that's a big challenge at times because the argument is, I don't think the future's
723
00:37:25.912 --> 00:37:26.492
been set yet.
724
00:37:26.612 --> 00:37:28.092
Yeah, exactly. Yeah.
725
00:37:28.272 --> 00:37:30.812
But if I were looking at it fairly,
726
00:37:31.612 --> 00:37:35.272
I think what I would say is, it kind of goes back to what we were talking about
727
00:37:35.312 --> 00:37:38.212
before, which is at the foundational layer,
728
00:37:39.112 --> 00:37:40.112
do you trust your data?
729
00:37:41.092 --> 00:37:45.062
Do you trust how it's used, stored, and applied,
730
00:37:45.092 --> 00:37:48.612
and approached? And if you have that as your foundational piece,
731
00:37:48.972 --> 00:37:51.992
then it makes the other choices easier.
732
00:37:52.172 --> 00:37:52.312
Mm-hmm.
733
00:37:52.392 --> 00:37:53.412
If that's in disarray,
734
00:37:54.652 --> 00:37:58.572
I imagine I walk into my house, and everything's everywhere because the kids have
735
00:37:58.612 --> 00:38:01.352
thrown everything everywhere. The toys are all over the place.
736
00:38:01.632 --> 00:38:05.572
It just doesn't feel like it's a calm place that I can actually
737
00:38:06.912 --> 00:38:09.252
maybe get dinner ready or things of that nature.
738
00:38:09.492 --> 00:38:12.632
But if I come into a house that's well-organized, I just feel this sense of
739
00:38:12.732 --> 00:38:16.692
calmness and ability to actually, okay, I'm going to go make dinner now
740
00:38:16.732 --> 00:38:20.672
because everything else is set. So when I look at that, it's the same
741
00:38:20.712 --> 00:38:23.652
with our data. The data foundation needs to be really
742
00:38:23.692 --> 00:38:26.712
well-established. Everybody feels confident about what they are.
743
00:38:27.312 --> 00:38:31.042
And then I think the next layer on top of that is we actually are helping
744
00:38:31.132 --> 00:38:35.102
people understand fundamentally whether it was using the data
745
00:38:35.102 --> 00:38:38.432
for reporting or analytics or using it for AI, it really doesn't matter.
746
00:38:38.912 --> 00:38:42.472
Do our people understand the data, understand the sensitivity,
747
00:38:42.512 --> 00:38:46.392
understand how it should be used? And if we have that, then
748
00:38:46.832 --> 00:38:50.312
we can start layering in the AI components on top of that.
749
00:38:50.412 --> 00:38:54.152
Okay, so we understand the data. Everybody understands proper
750
00:38:54.312 --> 00:38:54.572
use,
751
00:38:55.612 --> 00:38:59.552
whether it's AI or not. So the next layer we put on is a sandbox, and it's got
752
00:38:59.592 --> 00:39:03.561
guardrails on it. And everybody, you can be in that sandbox, you can
753
00:39:03.592 --> 00:39:07.472
play in that sandbox, and you can do it safely with the confidence that even if you
754
00:39:07.512 --> 00:39:08.742
mess up, you didn't mess up.
755
00:39:09.272 --> 00:39:09.652
Yeah.
756
00:39:09.712 --> 00:39:12.762
And that's how I think the organization builds the
757
00:39:12.832 --> 00:39:16.192
confidence to actually move forward from there.
758
00:39:16.212 --> 00:39:19.932
Because I think part of the solution for organizations is
759
00:39:20.232 --> 00:39:23.852
almost a hybrid one, which is work with partners you can really trust to
760
00:39:23.972 --> 00:39:26.192
actually understand that and who have done this.
761
00:39:26.502 --> 00:39:30.412
And I know that's self-serving and a bit selfish, but if you're really dealing with
762
00:39:30.472 --> 00:39:33.532
partners that have done this before and have the credentials,
763
00:39:34.452 --> 00:39:38.401
actually it helps accelerate yours because you can almost
764
00:39:38.432 --> 00:39:42.362
draft off some of what they've done to make sure that you're doing
765
00:39:42.392 --> 00:39:46.382
similar things that allow you to build up then your own solutions of how
766
00:39:46.432 --> 00:39:49.822
you want to use AI within your business, not just with a partner
767
00:39:49.992 --> 00:39:52.752
like Dayforce, for example, could bring to the table.
768
00:39:52.832 --> 00:39:53.012
Yeah.
769
00:39:53.072 --> 00:39:55.022
Because more and more we're hearing about agents.
770
00:39:55.072 --> 00:39:58.672
You brought it up earlier, Chris, right? Well, are you using our agent?
771
00:39:58.832 --> 00:40:02.632
Because maybe you're using our agent in your environment, not even through the
772
00:40:02.692 --> 00:40:05.272
Dayforce system, combined with your agent.
773
00:40:05.302 --> 00:40:08.492
And I think to your point, that orchestration that you talked about,
774
00:40:08.992 --> 00:40:12.272
that what we would call system of engagement or system of
775
00:40:12.352 --> 00:40:15.992
outcomes becomes like you and I sharing a box of
776
00:40:16.032 --> 00:40:17.542
Lego, and you have a
777
00:40:18.612 --> 00:40:22.572
rectangle, you pass it to me, and I click on a square, and then I pass
778
00:40:22.592 --> 00:40:24.371
it back to you, and you click on a...
779
00:40:24.452 --> 00:40:27.912
Like you start building these systems that can evolve
780
00:40:28.212 --> 00:40:31.992
leveraging partners and yourself, but that foundation is key
781
00:40:32.212 --> 00:40:35.832
in being able to do that and bring that up in the right standard.
782
00:40:35.892 --> 00:40:39.192
Yeah. No, I love that. When you was mentioning about the confidence, I was like, by
783
00:40:39.252 --> 00:40:41.512
creating that sandbox, that governance, et cetera,
784
00:40:42.512 --> 00:40:45.792
you create the confidence, but through the sandbox then
785
00:40:46.112 --> 00:40:47.392
create the competence.
786
00:40:48.152 --> 00:40:48.212
Mm-hmm.
787
00:40:48.292 --> 00:40:48.512
Right?
788
00:40:49.512 --> 00:40:50.072
But by-
789
00:40:50.112 --> 00:40:53.692
And confidence, too, I think is what you said earlier, which I really like
790
00:40:53.732 --> 00:40:57.592
because that's so critical. I'm confident in what I'm doing.
791
00:40:57.872 --> 00:40:58.192
Wow.
792
00:40:58.442 --> 00:40:59.232
Yeah. I've got a safe space.
793
00:40:59.262 --> 00:40:59.632
That really is-
794
00:40:59.992 --> 00:41:00.212
Mm-hmm
795
00:41:00.512 --> 00:41:01.012
... important.
796
00:41:01.412 --> 00:41:04.932
Yeah. And like you said, even if the
797
00:41:05.772 --> 00:41:09.572
AI is evolving and changing, as long as you have that
798
00:41:09.612 --> 00:41:13.432
foundation to operate from, the decision making, the team that you've put
799
00:41:13.471 --> 00:41:15.502
together, the AI governance-
800
00:41:15.692 --> 00:41:15.721
Mm-hmm
801
00:41:15.721 --> 00:41:19.592
... the data, and more importantly, what is it actually we're trying to solve for?
802
00:41:19.672 --> 00:41:23.032
What is our goal, right? Because otherwise you can get pulled in a thousand
803
00:41:23.072 --> 00:41:24.012
different directions. You're like-
804
00:41:24.032 --> 00:41:24.041
Mm-hmm
805
00:41:24.041 --> 00:41:27.592
... if this is not aligned with this, then the answer is no, and we move on.
806
00:41:27.752 --> 00:41:28.672
Otherwise, you kind of just
807
00:41:29.532 --> 00:41:31.862
get caught up with the new shiny object and the new thing, right?
808
00:41:32.492 --> 00:41:34.052
What are we actually driving towards?
809
00:41:34.212 --> 00:41:37.132
And that operating with confidence is so critical.
810
00:41:37.212 --> 00:41:40.292
I've heard from other people, "Oh, David, governance is just going to slow us
811
00:41:40.332 --> 00:41:43.872
down." And I kind of laughed. We have a small
812
00:41:43.912 --> 00:41:47.712
governance team. It's not huge. The team that actually looks at every idea that
813
00:41:47.772 --> 00:41:51.752
comes through is a three-person team that knows they can go out to a couple of
814
00:41:51.792 --> 00:41:53.072
other experts as needed.
815
00:41:53.272 --> 00:41:53.532
Mm-hmm.
816
00:41:54.092 --> 00:41:57.912
I think our SLA is now under five business days to turn around
817
00:41:57.932 --> 00:41:58.592
any idea-
818
00:41:58.632 --> 00:41:58.642
Really?
819
00:41:58.642 --> 00:42:01.652
... and say, "Yep, we can go forward with it." Yeah, it's that quick.
820
00:42:01.752 --> 00:42:01.912
Wow.
821
00:42:02.052 --> 00:42:04.171
So it doesn't have to be slow.
822
00:42:05.072 --> 00:42:08.102
And you do hit some circumstances where you need back and forth with, let's say, a
823
00:42:08.192 --> 00:42:10.562
partner or a vendor, but it can move quickly.
824
00:42:10.612 --> 00:42:13.782
It actually enables the business to move faster.
825
00:42:13.812 --> 00:42:15.862
It doesn't cause the business to move slower.
826
00:42:16.292 --> 00:42:18.832
Yeah. Listen, I feel like I could talk to you forever, but I've got to let you go
827
00:42:18.872 --> 00:42:19.692
at some point.
828
00:42:20.842 --> 00:42:21.352
Fair enough.
829
00:42:21.412 --> 00:42:25.192
Before I let you go, we covered a lot, but what would be your sort of
830
00:42:25.252 --> 00:42:28.572
parting advice for those HR leaders that are listening right now that
831
00:42:29.372 --> 00:42:32.892
everyone's on their journey? And then also, where's the best place for people
832
00:42:32.952 --> 00:42:36.272
to connect with you if they want to reach out and say hi and learn more?
833
00:42:38.352 --> 00:42:38.972
Definitely
834
00:42:39.852 --> 00:42:43.512
starting with that, the simplest part, for biases of recency first-
835
00:42:44.468 --> 00:42:46.388
Dayforce.com is a great place to start with.
836
00:42:46.468 --> 00:42:49.488
But if you have questions, one of the things I think we've really prided ourselves
837
00:42:49.548 --> 00:42:52.728
on working with everyone is, reach out to me on LinkedIn,
838
00:42:52.768 --> 00:42:56.668
connect. And I may not be able to answer your question, but I can connect you
839
00:42:56.688 --> 00:42:57.648
with the people that can.
840
00:42:58.588 --> 00:43:02.408
We put out, I think, a lot of great content that is about helping teach, not
841
00:43:02.468 --> 00:43:03.208
necessarily
842
00:43:04.228 --> 00:43:06.868
buy our product, but this is how we do it.
843
00:43:06.928 --> 00:43:09.708
So I think there's a lot out there. Please follow me.
844
00:43:09.788 --> 00:43:13.008
Happy to do that as well to support that learning journey.
845
00:43:13.608 --> 00:43:15.728
And make sure you have the basics in place.
846
00:43:15.988 --> 00:43:19.958
If you don't have a governance, an idea of how you want to do governance, if
847
00:43:20.008 --> 00:43:23.468
you don't have any idea how you actually want to create a sandbox
848
00:43:23.848 --> 00:43:24.468
environment,
849
00:43:25.548 --> 00:43:28.728
get those things done first. That's like 101.
850
00:43:28.968 --> 00:43:32.688
You need those things in place. And organizations are moving quickly that way, but
851
00:43:33.288 --> 00:43:37.248
that I think has to be there. And then decide really
852
00:43:37.308 --> 00:43:39.928
through that governance process what you want to tackle first.
853
00:43:40.288 --> 00:43:43.948
Don't be everything everywhere all at once, and don't listen to the hype
854
00:43:44.028 --> 00:43:47.668
cycle that's outside there telling you that you're not yet
855
00:43:47.708 --> 00:43:51.648
agentic. Go through that process of learning, and make
856
00:43:51.728 --> 00:43:53.608
sure that you take the organization with you.
857
00:43:53.788 --> 00:43:57.648
And I think one of my colleagues put it best when I think of all the
858
00:43:57.728 --> 00:43:58.688
flying that we do.
859
00:43:59.588 --> 00:44:00.308
The HR leader,
860
00:44:01.388 --> 00:44:05.128
put your oxygen mask on first. Make sure you've taken care of yourself
861
00:44:05.228 --> 00:44:09.108
first before you help others. And I think that's another really great
862
00:44:09.168 --> 00:44:12.288
lesson learned, because you're going to be more powerful when you have a certain
863
00:44:12.348 --> 00:44:15.968
level of confidence in how to apply these things and what it means to the
864
00:44:16.008 --> 00:44:18.028
business. And then finally, partner.
865
00:44:18.548 --> 00:44:22.168
Your relationship with your CIO. If you have a chief privacy
866
00:44:22.228 --> 00:44:25.848
officer or someone who looks at those areas and may be on your legal
867
00:44:25.968 --> 00:44:29.628
side, work with them closely and start aligning those
868
00:44:29.788 --> 00:44:33.068
interests so that you're not going in many different directions.
869
00:44:33.148 --> 00:44:36.428
And those would be some of the basics I would say right off the bat.
870
00:44:36.448 --> 00:44:40.268
And then my self-serving one is work with a trusted partner, one that's
871
00:44:40.348 --> 00:44:44.008
demonstrated that they actually know what they're doing, that they have the right
872
00:44:44.048 --> 00:44:48.028
certifications and attention to detail so that you can
873
00:44:48.068 --> 00:44:49.858
draft in behind what they're doing.
874
00:44:50.248 --> 00:44:54.208
Because there's incredible data that we share with our customers about our Dayforce
875
00:44:54.268 --> 00:44:54.728
Assistant.
876
00:44:55.148 --> 00:44:55.158
Mm-hmm.
877
00:44:55.488 --> 00:44:58.848
Why is 18% of all questions around time away from work and 10%
878
00:44:58.888 --> 00:45:02.548
around payroll? And how you can use that data to
879
00:45:02.628 --> 00:45:06.608
actually drive great AI experiences, to your point, employee experiences
880
00:45:06.668 --> 00:45:07.008
as well.
881
00:45:07.408 --> 00:45:07.488
Yeah.
882
00:45:07.518 --> 00:45:11.338
So those are some of the things I would consider when I was walking through that.
883
00:45:11.768 --> 00:45:15.458
Amazing, Dave. Well, I've really enjoyed the conversation, and congrats to you and
884
00:45:15.488 --> 00:45:18.608
the team on the journey so far. And when our
885
00:45:19.228 --> 00:45:21.928
audience does call in, now they know the questions to ask your team.
886
00:45:23.288 --> 00:45:27.228
Fair enough. We're ready for them, Chris, so I think that's exciting for me.
887
00:45:27.297 --> 00:45:28.968
That'll be the easiest part of our day.
888
00:45:29.108 --> 00:45:31.608
That'll be your real test
889
00:45:32.428 --> 00:45:32.928
to the team.
890
00:45:32.948 --> 00:45:36.288
Yes. I think that's a great one to look at. Well, you have a great day.
891
00:45:36.468 --> 00:45:37.228
You too. Thank you.
David Lloyd, SVP Platform Engineering & Chief AI Officer at Dayforce.