Shownotes
In this special live episode of the Future Proof HR podcast, recorded on the floor at Transform 2026 in Las Vegas, Thomas Kunjappu sits down with Conor Sweeney, VP of People at Form Health, for a quick but packed conversation about what AI in HR actually looks like in practice at a fast-growing telehealth company.
Form Health is the national leader in science-backed obesity care management and cardiometabolic health, operating across 30+ states in the United States, and is a company scaling fast with two distinct employee populations: clinical and corporate. Conor and his team have had to think creatively about how to make it easier for people to do their jobs every day.
They get into what it really looks like to implement agentic AI for knowledge management, the unglamorous work of keeping that content accurate enough to actually be useful, and what's next for AI in talent acquisition. Conor also shares what he's taking away from Transform: HR leaders aren't just adopting AI anymore, they're owning it.
A candid, grounded conversation straight from one of HR's biggest stages.
Topics Discussed:
- Using agentic AI for knowledge management and why Guru has worked for Form Health
- The real challenge of content accuracy when feeding an AI system
- How AI can help TA teams spend less time screening and more time advising
- HR's growing ownership of AI strategy across industries
- Moving from reactive to proactive in compliance and HR ops
Additional Resources:
Transcripts
We're live here at Transform in Las Vegas and we are talking about
2
00:00:08,809 --> 00:00:10,399
Future Proofing HR.
3
00:00:10,399 --> 00:00:12,360
As always, I'm your host, Thomas Kunjappu.
4
00:00:12,379 --> 00:00:17,630
And I'm here with Conor Sweeney, who
is the VP of People at Form Health.
5
00:00:17,680 --> 00:00:19,510
Tell us a little bit about
Form Health before we talk
6
00:00:19,510 --> 00:00:21,100
more about Future Proofing HR.
7
00:00:21,160 --> 00:00:22,210
Connor: Yeah, absolutely.
8
00:00:22,210 --> 00:00:24,459
I've been with Form Health
for close to a year now.
9
00:00:24,799 --> 00:00:28,849
we are the national leader in
science-backed obesity, care
10
00:00:28,849 --> 00:00:30,319
management and Cardiometabolic health.
11
00:00:30,319 --> 00:00:34,090
So we are telehealth practice that
operates here in the United States.
12
00:00:34,450 --> 00:00:35,050
Thomas: That's amazing.
13
00:00:35,099 --> 00:00:38,054
Tell us a little bit about the
AI journey at your organization.
14
00:00:38,785 --> 00:00:39,954
Connor: There's a couple
different paths, right?
15
00:00:39,954 --> 00:00:42,654
And I think we've been talking a lot
about that this week where there is
16
00:00:42,654 --> 00:00:46,014
the what is the business doing from
a business model perspective, right?
17
00:00:46,014 --> 00:00:50,794
Around call it outward facing
technology, or in our case how our
18
00:00:50,855 --> 00:00:53,615
clinicians are engaging with the
tools and systems that they're using
19
00:00:53,615 --> 00:00:55,324
to provide care to their patients.
20
00:00:55,805 --> 00:00:57,995
And then the other track is what
are we doing to just make it
21
00:00:57,995 --> 00:00:59,495
easier for our employees to work?
22
00:00:59,794 --> 00:01:00,065
Thomas: Right?
23
00:01:00,185 --> 00:01:02,499
Connor: And for me, what's top of
mind for me is that second bucket.
24
00:01:03,449 --> 00:01:05,339
Which is, how do we just
make it easier to work here?
25
00:01:05,468 --> 00:01:08,339
One of the things that's been really
great over the last six months is
26
00:01:08,339 --> 00:01:12,729
we've really leaned into Agentic AI for
the purposes of knowledge management.
27
00:01:12,729 --> 00:01:17,610
So we implemented a system that you could
equate to call it an intranet, right?
28
00:01:17,660 --> 00:01:19,100
Something that we're
all very familiar with.
29
00:01:19,430 --> 00:01:23,870
But something that is far more advanced
where folks in our organization
30
00:01:23,870 --> 00:01:28,320
can go into this platform, type in
a question, and hopefully whatever
31
00:01:28,320 --> 00:01:30,280
they're looking for is provided to them.
32
00:01:30,710 --> 00:01:33,560
We use Guru as our platform and
it's been really successful.
33
00:01:33,560 --> 00:01:36,740
In healthcare there's a lot of
different policies and a lot
34
00:01:36,740 --> 00:01:40,530
of different intricacies to
ways of working and process.
35
00:01:40,830 --> 00:01:44,550
And so being able to provide employees
with a platform where they don't have
36
00:01:44,550 --> 00:01:48,360
to spend minutes to hour searching
through folders for things they can
37
00:01:48,360 --> 00:01:51,210
just ask the question and hopefully
get the result they're looking for.
38
00:01:51,259 --> 00:01:52,809
Has been really successful for us so far.
39
00:01:53,544 --> 00:01:56,774
Thomas: What was the impetus of working
on this particular project to begin with?
40
00:01:56,774 --> 00:02:00,004
What was like the the pain or
like the idea to prioritize this?
41
00:02:00,054 --> 00:02:02,544
Connor: I can't take credit because
when I stepped into the organization,
42
00:02:02,544 --> 00:02:05,874
our product team was playing a huge
role in putting this into place.
43
00:02:05,874 --> 00:02:12,204
But the impetus was really around we
are growing fast and we have really
44
00:02:12,204 --> 00:02:14,484
two different employee populations.
45
00:02:14,484 --> 00:02:17,304
We have folks that are patient
facing our clinicians that are
46
00:02:17,304 --> 00:02:18,954
really frontline every single day.
47
00:02:19,374 --> 00:02:21,644
And then we have the folks that
support the business, the G&A
48
00:02:21,774 --> 00:02:23,304
functions, marketing, commercial.
49
00:02:23,304 --> 00:02:26,394
The needs for those employees around
knowledge management are very different.
50
00:02:26,844 --> 00:02:30,514
And so when you think about just
slapping Google folders into a
51
00:02:30,524 --> 00:02:32,524
GDrive and saying best of luck.
52
00:02:32,524 --> 00:02:34,084
Go find what you need to look for.
53
00:02:34,534 --> 00:02:35,074
It's tough.
54
00:02:35,074 --> 00:02:36,864
You need to have something
that's more curated
55
00:02:36,864 --> 00:02:40,594
to what folks need within the context of
the work they're doing every single day.
56
00:02:41,054 --> 00:02:42,374
It was an efficiency play.
57
00:02:42,374 --> 00:02:44,174
It was a how do we make
it easier to work here.
58
00:02:44,174 --> 00:02:49,554
And also just how do we continue to lean
into Agentic AI opportunities to aid
59
00:02:49,554 --> 00:02:51,094
in our growth trajectory as a company.
60
00:02:51,574 --> 00:02:52,444
Thomas: I love that logic.
61
00:02:52,494 --> 00:02:55,424
Tell me a little bit about any of the
challenges or things that you learned
62
00:02:55,424 --> 00:03:00,894
on the execution of this project that
you had to overcome to get to the I
63
00:03:00,894 --> 00:03:03,134
don't know if the promised land is
you've arrived at it or maybe still
64
00:03:03,134 --> 00:03:04,304
elusive or it's an ongoing journey.
65
00:03:04,304 --> 00:03:07,724
But what have you learned so far
through the execution of this project?
66
00:03:07,774 --> 00:03:11,254
Connor: Like any cross-functional
project, it's the stakeholder
67
00:03:11,254 --> 00:03:15,064
management, stakeholder engagement,
setting expectations for what teams
68
00:03:15,064 --> 00:03:16,444
need to play, which role and what.
69
00:03:16,904 --> 00:03:19,754
When you're talking AI focused
knowledge management, there's a lot of
70
00:03:19,754 --> 00:03:23,504
governance that needs to be put in place
around is the information up to date?
71
00:03:23,744 --> 00:03:26,904
Are we feeding the system with
the right level of information?
72
00:03:27,224 --> 00:03:31,049
Or in many cases when we have folks
searching for things, it may not exist.
73
00:03:31,289 --> 00:03:35,159
So how do we drive accountability
on creating that SOP, that workflow,
74
00:03:35,159 --> 00:03:39,189
that policy, and how quickly can we
do it in a way that's responsible
75
00:03:39,189 --> 00:03:40,504
and gets folks what they need?
76
00:03:41,164 --> 00:03:44,434
I think the biggest challenge
is making sure that the accuracy
77
00:03:44,434 --> 00:03:47,194
of the content that lives in the
system is what it needs to be.
78
00:03:47,504 --> 00:03:51,043
We need the content to be
accurate in order for the agent
79
00:03:51,043 --> 00:03:52,394
to provide the information.
80
00:03:52,394 --> 00:03:53,414
But that takes time.
81
00:03:53,414 --> 00:03:56,984
It takes folks across every business
unit of the company to step in and
82
00:03:56,984 --> 00:04:01,574
say, yeah, of the dozens or hundreds
of policies or insight docs that we
83
00:04:01,574 --> 00:04:04,734
own, what's our cadence for making
sure that this stuff is up to date?
84
00:04:05,024 --> 00:04:05,294
Thomas: Yeah.
85
00:04:05,294 --> 00:04:08,414
It starts to expose I guess
organizational debt a little bit, right?
86
00:04:08,414 --> 00:04:08,894
About a hundred percent.
87
00:04:08,944 --> 00:04:09,904
With documentation.
88
00:04:09,964 --> 00:04:12,404
And it actually feels this
is a path people go down.
89
00:04:12,434 --> 00:04:16,394
'Cause you think about Agentic AI and
these digital workers is doing things
90
00:04:16,394 --> 00:04:18,164
for you, but then what feeds that?
91
00:04:18,194 --> 00:04:20,803
It's processes or documents.
92
00:04:20,803 --> 00:04:23,594
In this case, there's something
like underlying that which needs
93
00:04:23,594 --> 00:04:25,484
to be taken care of by the company.
94
00:04:25,484 --> 00:04:29,604
And it's exposes how well you
understand what it is that you wanna do.
95
00:04:29,604 --> 00:04:34,144
Because if you can't have a for example,
a new HR generalist or I don't know,
96
00:04:34,144 --> 00:04:38,214
if a product specialist is supposed to
help the operations team get trained
97
00:04:38,214 --> 00:04:39,744
up quickly on answering questions.
98
00:04:39,864 --> 00:04:43,144
How can you expect the AI to as well?
99
00:04:43,304 --> 00:04:46,334
It sounds like the big challenge
that was kinda like making sure the
100
00:04:46,364 --> 00:04:48,854
inputs for the AI were like accurate.
101
00:04:49,044 --> 00:04:51,794
And did you find it was any different if
you're documenting things for employees
102
00:04:51,794 --> 00:04:53,939
to look it up themselves versus for AI?
103
00:04:54,519 --> 00:04:57,609
Connor: Not necessarily, but I think at
the end of the day, it's having to work
104
00:04:57,609 --> 00:04:59,319
cross-functionally to create process.
105
00:04:59,679 --> 00:05:02,919
And I think, driving accountability
when there's multiple owners of
106
00:05:02,919 --> 00:05:04,239
something can be really difficult.
107
00:05:04,579 --> 00:05:05,149
We've lucked out.
108
00:05:05,149 --> 00:05:08,029
We have everybody that I work with
is the most amazing human on earth.
109
00:05:08,079 --> 00:05:09,999
it's not like it's a
particular challenge for us.
110
00:05:09,999 --> 00:05:12,999
But when you think about all the
things that come with growing a
111
00:05:12,999 --> 00:05:15,559
company that's moving really fast.
112
00:05:15,719 --> 00:05:17,529
Prioritization can be a challenge.
113
00:05:17,909 --> 00:05:20,309
I think the other thing that can be
challenging too is we're bringing a
114
00:05:20,309 --> 00:05:23,609
lot of folks on board over the course
of the months and years ahead, right?
115
00:05:23,609 --> 00:05:27,779
As our kind of company is growing, we need
to bring more people into the ecosystem.
116
00:05:28,049 --> 00:05:31,739
And these folks are coming from
oftentimes more mature, larger,
117
00:05:31,739 --> 00:05:33,089
more built out backgrounds, right?
118
00:05:33,189 --> 00:05:36,519
So their expectations of what
documentation looks like, what
119
00:05:36,519 --> 00:05:39,689
workflows look like is becoming
increasingly more And so what
120
00:05:39,689 --> 00:05:41,129
can we do to keep up with that?
121
00:05:41,289 --> 00:05:43,359
And I think that's for me and
other folks on my team, having
122
00:05:43,359 --> 00:05:46,389
backgrounds in larger company
environments has been really helpful.
123
00:05:47,169 --> 00:05:47,709
Thomas: That's awesome.
124
00:05:47,709 --> 00:05:49,819
So scaling up like in this environment.
125
00:05:49,869 --> 00:05:51,189
Let me ask you about the future then.
126
00:05:51,199 --> 00:05:52,939
So you've kinda worked on
this project, what other
127
00:05:53,299 --> 00:05:56,739
kind of things are you looking at next
as you're thinking about growth and
128
00:05:56,789 --> 00:05:58,439
like investment in this direction?
129
00:05:58,569 --> 00:06:00,679
Connor: Shout out to my
talent acquisition team.
130
00:06:00,679 --> 00:06:03,439
Because a big search for me here
this week is really thinking
131
00:06:03,439 --> 00:06:05,779
about how we use AI to manage
132
00:06:06,049 --> 00:06:10,299
top of the funnel optimization and
how do we give our strategic talent
133
00:06:10,299 --> 00:06:12,759
partners more time to do strategic work?
134
00:06:13,129 --> 00:06:16,909
Being a telehealth company and being
one that's growing really fast,
135
00:06:16,919 --> 00:06:21,219
we've got a lot of folks, doctors,
APPs, registered dieticians that are
136
00:06:21,219 --> 00:06:22,839
applying to work in our organization.
137
00:06:23,289 --> 00:06:26,289
Not a bad problem to have, but
when you think about the amount of
138
00:06:26,649 --> 00:06:30,389
documentation that our TA team needs
to look for to ensure that we don't
139
00:06:30,389 --> 00:06:33,539
have bots applying for roles to ensure
that the licensing and credentialing
140
00:06:33,539 --> 00:06:35,039
expectations are where they need to be.
141
00:06:35,089 --> 00:06:38,149
Thousands and thousands of
resumes is a lot for a small
142
00:06:38,149 --> 00:06:39,409
but mighty team to go through.
143
00:06:39,409 --> 00:06:41,629
I think what's been great about this
week is seeing that there's a lot of
144
00:06:41,629 --> 00:06:46,669
really good momentum around how AI can
play a role in helping to manage the
145
00:06:46,669 --> 00:06:48,679
optimization of like top of the funnel.
146
00:06:49,349 --> 00:06:53,609
How we get TA teams spending less time
screening and kind of more time advising
147
00:06:53,609 --> 00:06:55,229
and working with hiring managers.
148
00:06:55,639 --> 00:06:56,929
I'm really looking forward to that.
149
00:06:57,649 --> 00:06:59,058
Thomas: That's a great, like a use case.
150
00:06:59,058 --> 00:07:00,739
Glad you found some
potential partners here.
151
00:07:00,789 --> 00:07:03,219
Are there any other major takeaways
for you from the conference so
152
00:07:03,219 --> 00:07:06,639
far or like thematically things
that you're taking back to your
153
00:07:06,639 --> 00:07:08,139
team as you walk away from Vegas?
154
00:07:08,449 --> 00:07:13,039
Connor: I'm so proud of what HR
leaders across all the industries.
155
00:07:13,069 --> 00:07:16,104
Like how we are working with
our organizations on what is
156
00:07:16,104 --> 00:07:18,644
a rapidly evolving situation.
157
00:07:19,074 --> 00:07:23,039
If I think about just even being here last
year or the year before, the conversations
158
00:07:23,039 --> 00:07:27,569
that we were having around adopting AI
and kind of infusing later-stage kind of
159
00:07:27,569 --> 00:07:31,379
technical innovation into our companies,
it was very different than today.
160
00:07:31,379 --> 00:07:34,919
Me being able to sit down with
a group of heads of HR or CHROs
161
00:07:34,919 --> 00:07:38,159
and they're like, yep, I now own
this aspect of our AI strategy.
162
00:07:38,159 --> 00:07:42,599
Or, hey, my Head of People Ops is
now fully in control of this piece.
163
00:07:42,599 --> 00:07:46,099
And so I'm really happy to see that
there's emphasis on leaning into people
164
00:07:46,099 --> 00:07:49,399
teams as that strategic partner to kind
of bring some of this technological
165
00:07:49,399 --> 00:07:52,804
evolution to life to make it easy
for our employees to do their job.
166
00:07:53,874 --> 00:07:58,764
I'm proud of what kind of cross
industry wise we've empowered HR leaders
167
00:07:58,764 --> 00:08:00,324
to do over the last couple years.
168
00:08:00,444 --> 00:08:00,864
Thomas: I love that.
169
00:08:00,864 --> 00:08:02,984
And you're starting to see some of
those results come in terms of the
170
00:08:02,984 --> 00:08:04,834
ownership within the organization.
171
00:08:04,834 --> 00:08:09,214
So last question, we always
talk about Future Proofing HR.
172
00:08:09,219 --> 00:08:12,374
Maybe you're getting to that a little
bit with this last answer, but how
173
00:08:12,374 --> 00:08:14,894
do you think about within your own
organization, future proofing the
174
00:08:14,894 --> 00:08:19,174
HR function and ensuring that you're
relevant and thriving in the age of AI?
175
00:08:20,114 --> 00:08:20,504
Connor: Yeah.
176
00:08:20,504 --> 00:08:23,894
I mean, look, I would say we're off
to a good start, but like any company,
177
00:08:23,894 --> 00:08:25,994
there's a lot of opportunity for us.
178
00:08:26,184 --> 00:08:30,334
I talked a little bit earlier about
the way to optimize TA functionality.
179
00:08:30,534 --> 00:08:34,703
And the intersection of how that
can be aided with the AI technology.
180
00:08:34,764 --> 00:08:37,774
I'd say the second piece is look as a
healthcare organization, there's a lot
181
00:08:37,774 --> 00:08:39,688
of aspects of compliance and kind of
182
00:08:40,279 --> 00:08:43,779
regulatory implications that
are always top of mind, right?
183
00:08:43,779 --> 00:08:44,889
Things are always changing.
184
00:08:44,889 --> 00:08:50,289
And so if you think about the role of
an HR ops leader or a payroll leader,
185
00:08:50,649 --> 00:08:54,729
folks within my own team that are just
trying to keep up with the expectations
186
00:08:54,759 --> 00:09:00,474
of wow we have employees in 30+ states
and the regulations around how employees
187
00:09:00,474 --> 00:09:03,354
get paid or the different trainings
they have to take or the licensing
188
00:09:03,354 --> 00:09:05,774
and credentialing processes that
are applicable to those states, like
189
00:09:05,864 --> 00:09:07,104
that stuff changes a lot.
190
00:09:07,464 --> 00:09:11,054
And so I think there's a tremendous
opportunity to lean into technology
191
00:09:11,054 --> 00:09:13,874
and AI to help us have a better
understanding of what's ahead.
192
00:09:14,444 --> 00:09:17,094
I think people teams for decades have
been plagued with like reactiveness.
193
00:09:18,309 --> 00:09:18,759
Thomas: Absolutely.
194
00:09:18,829 --> 00:09:21,919
Connor: I think AI gives us
potentially a leg up on how can
195
00:09:21,919 --> 00:09:23,329
we be more proactive, right?
196
00:09:23,329 --> 00:09:27,289
So not us finding out and this is made
up example, but not us finding out,
197
00:09:27,499 --> 00:09:31,419
hey, wow we've been non-compliant with
these three or four things over the
198
00:09:31,419 --> 00:09:32,919
last 12 months and we had no idea.
199
00:09:32,919 --> 00:09:36,359
But more so hey, if we don't do this
now 18 months from now, we're gonna
200
00:09:36,359 --> 00:09:37,559
run into some compliance issues.
201
00:09:37,559 --> 00:09:39,929
And I think there's some
really cool technology out
202
00:09:39,929 --> 00:09:40,829
there that can help with that.
203
00:09:41,469 --> 00:09:43,779
Thomas: When you say reactive to
proactive, what comes to mind for
204
00:09:43,779 --> 00:09:47,694
me is more the burnt out HR like
payroll specialists or generalists
205
00:09:47,694 --> 00:09:52,614
who is like reacting to like tickets
or their work queue constantly.
206
00:09:52,614 --> 00:09:55,584
And so they can't actually
focus on the thing that could be
207
00:09:55,584 --> 00:09:57,864
strategically adding value, like
absolutely 18 months down the line.
208
00:09:57,864 --> 00:10:02,524
So yeah, big opportunity there to
upskill, reskill and leverage AI to
209
00:10:02,584 --> 00:10:03,994
change what people are doing every day.
210
00:10:04,054 --> 00:10:08,124
I love hear these insights and
the idea that HR leaders are
211
00:10:08,124 --> 00:10:12,534
really leading in playing a key
role in AI transformations within
212
00:10:12,534 --> 00:10:15,581
organizations, including at Form Health
213
00:10:15,705 --> 00:10:16,245
. Connor: That's right.
214
00:10:16,355 --> 00:10:19,805
Thomas: How can folks follow your
journey or stay connected with you Conor?
215
00:10:20,515 --> 00:10:21,010
Connor: I love LinkedIn.
216
00:10:22,255 --> 00:10:25,525
It's like I spend more time on LinkedIn
than any other social media platform.
217
00:10:25,575 --> 00:10:28,935
and it's not just because I've got
lots of people that I greatly respect
218
00:10:29,235 --> 00:10:30,945
that I think share all the insights.
219
00:10:30,945 --> 00:10:37,265
But to me it's where a lot of really
good information is being shared.
220
00:10:37,615 --> 00:10:39,595
I also just selfishly having
been doing this for a while.
221
00:10:39,595 --> 00:10:41,875
Love to see the progress of folks
that I've worked with in the past.
222
00:10:41,875 --> 00:10:43,505
Please, feel free to add me on LinkedIn.
223
00:10:43,535 --> 00:10:46,615
Always happy to chat,
coffee, wine, whatever it is.
224
00:10:46,715 --> 00:10:47,115
Thomas: Wine today...
225
00:10:47,115 --> 00:10:47,495
Connor: But that's wine today.
226
00:10:47,870 --> 00:10:48,920
Might be coffee tomorrow.
227
00:10:48,920 --> 00:10:49,160
There you go.
228
00:10:49,160 --> 00:10:50,110
Definitely will be coffee tomorrow.
229
00:10:51,500 --> 00:10:54,750
Thomas: So great connecting with
you Conor and great talking about
230
00:10:54,800 --> 00:10:57,940
what you've been doing there with
AI initiatives at Form Health.
231
00:10:58,000 --> 00:11:01,490
And for everyone out there, this
has been another micro episode of
232
00:11:01,610 --> 00:11:05,990
Future Proof HR as we figure out a
way to all thrive in the age of AI.
233
00:11:06,740 --> 00:11:07,430
Thank you so much.
234
00:11:08,580 --> 00:11:11,640
Thomas Kunjappu: Thanks for joining
us on this episode of Future Proof HR.
235
00:11:11,880 --> 00:11:14,490
If you like the discussion, make
sure you leave us a five star
236
00:11:14,490 --> 00:11:17,370
review on the platform you're
listening to or watching us on.
237
00:11:17,790 --> 00:11:20,730
Or share this with a friend or colleague
who may find value in the message.
238
00:11:21,030 --> 00:11:25,980
See you next time as we keep our pulse on
how we can all thrive in the age of AI.
