December 5, 2025 • 26 mins
What actually changes when AI agents become part of your workforce — and which human skills still matter most?
In this episode, host Andreas Welsch talks with HR and talent-intelligence veteran Todd Raphael about the practical realities of bringing agentic AI into organizations. They move beyond proofs-of-concept to ask the tough questions: How do agents fit into daily workflows, what invisible human contributions should you protect, and how should HR and IT collaborate to redesign roles, org charts, and the employee lifecycle?
Listen for concrete thinking and strategic framing, including:
- The hidden value humans bring: Why many critical contributions (trusted relationships, customer touchpoints, institutional memory) don’t appear on job descriptions — and what that means when you automate tasks.
- Rethinking structure and advancement: How flatter org models and new measures of impact (knowledge, networks, influence) may change who gets promoted and how leadership is defined.
- HR’s seat at the table: Why HR is uniquely positioned to plan holistically for hire-to-retire changes, from skills-based hiring to internal marketplaces, reskilling, and retention when agents handle more tasks.
You’ll also hear examples and practical prompts for leaders: identify the intangible work that must remain human, map tasks vs. relationships before automating, and start workforce planning that considers people and agents together.
If you’re an HR leader, people manager, or technology decision-maker trying to turn agent hype into durable business outcomes, this episode gives you a playbook to start redesigning work the right way.
Tune in now to learn how to protect human advantage and build an effective human+agent workforce.
Questions or suggestions? Send me a Text Message.
***********
Disclaimer: Views are the participants’ own and do not represent those of any participant’s past, present, or future employers. Participation in this event is independent of any potential business relationship (past, present, or future) between the participants or between their employers.
Level up your AI Leadership game with the AI Leadership Handbook:
https://www.aileadershiphandbook.com
More details:
https://www.intelligence-briefing.com
All episodes:
https://www.intelligence-briefing.com/podcast
Get a weekly thought-provoking post in your inbox:
https://www.intelligence-briefing.com/newsletter
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Andreas Welsch (00:31):
Hey, it's the end of 2025 and things are
getting real.
What actually changes when youbring agents into a business?
How do teams need to be set upand what do HR and business
leaders need to get we'll talkabout all that.
So let's jump right in.
Welcome to What's the BUZZ?,where leaders share how they
have turned hype into outcome.
(00:52):
Today, we'll talk about how todesign your workforce for
agentic AI and who better totalk about it than someone who's
actively working in this space.
Todd, Raphael.
Todd, thank you so much forjoining.
Todd Raphael (01:02):
Thanks for having me.
Andreas Welsch (01:03):
Hey, Todd, maybe not everybody in the audience is
familiar with you or has comeacross your content.
I certainly have, and I've beena big fan of the things that you
put out for a long time.
Why don't you tell us a littlebit about yourself, who you are
and what you do?
Todd Raphael (01:15):
No problem.
I spent more than a coupledecades at a couple
organizations Cranes and thenERE on conferences and
publications for recruiters, HRprofessionals and others.
And then went to a HR technologycompany called Eightfold fairly
early on and helped grow thatmarketing department, grow the
organization dramatically overthree years.
(01:37):
And then was recruited to go toa competing organization called
Sky Hive, which sold toCornerstone.
Cornerstone on demand.
About a couple years ago, andI've been working as a helping
HR tech companies, talentacquisition technology companies
and related companies, staffingcompanies and so on over the
last couple years, help figureout how to differentiate
(02:00):
themselves, how to communicatewhat they do, how to explain
what they do communicate, createcontent, work with influencers
and that kind of thing.
Help them figure themselves outand how to find themselves and
communicate who they are.
Andreas Welsch (02:12):
Thank you for sharing.
I feel a lot of times marketinggets a bit of a bad reputation
and I've previously been in, inmarketing myself is in my
career.
But I feel you also, get a verygood grasp on what's actually
happening in, in the market.
What are different vendorsdoing, what do buyers actually
look for, which is why I'm soexcited to have you on the show
and learn from your perspectiveas, as well what you're seeing.
(02:35):
Todd, it's the end of 2025.
It feels like we've talked aboutAI agents as an industry for, I
don't know a year or so.
It feels like it's already beena decade in AI years.
And it feels like now we'rereally at this point where we're
seeing more and moreorganizations, more and more
leaders, more and more HRleaders also come to the table
(02:55):
and say what do we do?
We're spinning up individualproofs of concepts and pilots
and we're working with differentvendors to try things out, and I
think that's great.
Finally we're moving to towardsaction, but when I talk to
leaders, when I attend vendorconferences and it's probably
similar for you as well, itseems that everybody just talks
(03:17):
about the building part ofagents, but there's not a lot of
talk about what actually happensafterward.
What happens when you bring themto the organization?
What happens when they are inproduction?
I'm curious with all this talkabout did your human employees
get agent employees?
Do we now need to draw parallelsto how we manage humans as well?
Can we apply some of that toagents?
(03:38):
What are you seeing?
Todd Raphael (03:39):
Yeah, I've been to a bunch of conferences this last
couple years and including tworecently.
And there was all kinds ofbooths having to do with agents
and buying agents.
And one of'em, you get to buildyour own agents right at the
booth there on a little iPad,and then after five minutes
you're done, you got your agent.
But yeah, there was not a wholelot of explanation exactly.
What does your day look like?
What is it?
How is this agent kind of workin your workflow?
(04:02):
Like you get up, you have a job,you find out there's someone
quit, so you have an open job.
How do you fill that job usingan agent?
What do you do?
Like step A, step B, step C,step D and E.
Like, where's the agent fittingin and what do they do and what
do they not do?
And then once that agent's partof your workflow, your hiring
process.
(04:22):
What then do you do?
How do you know in six months ifit's biased or how do you know
if it's accurate?
It may be perfectly fair, butnot very accurate.
You may be not really an imp,maybe not be an improvement over
how things were without an agentor maybe there's some
personalization with the agentthat's no longer working very
effectively or something.
But, so those are, I think, openquestions, but certainly there's
(04:43):
not a lot of informationexactly.
About how do you spend a day orspend a week or a month.
With agents, what exactly, howdoes that work in your life,
your work life, your job?
Andreas Welsch (04:52):
Yeah.
I mean I've been creating somecourses with LinkedIn Learning
on the topic what happens whenyou bring agents to business.
And some of them came out at theend of last year.
So even very early in, in theirwave.
But I'm sure things have beenevolving and especially as you
and your businesses are thinkingabout this too.
My, my recommendation would be,think about it.
How, yeah.
(05:13):
How do we use these things once,once they're in our business in
that sense.
And I feel with all that focuson building, to your point,
there's not enough conversationabout what happens afterwards.
Yeah.
And that, that opens, manyparallels in my point of view to
how do we work with people,right?
We wanna make sure thatdepending on the role or the
(05:33):
company that we run, backgroundchecks are you who you say you
are.
Permissions for your role.
You get certain permissions tocertain systems, but not more
extensive than what you reallyneed to guidelines.
Codes of conduct.
I want you to behave in acertain way.
I'm thinking about what iftechnology teams build agents in
every different department orevery user builds it slightly
(05:54):
differently.
Do we need to ground them insome sort of code of conduct,
for example?
What do you think?
Todd Raphael (06:00):
I think that there's this idea that, so a lot
of companies I work with areworking on talent intelligence
or skills intelligence, andthey're taking AI and they're
taking jobs and trying to figureout what tasks and skills are
involved in each job.
That by breaking a job up intotasks and skills, you can help
(06:21):
predict the future.
You can say, oh, this job, halfof it may no longer be necessary
to be done by a human, or thesetwo jobs could be melded
together in the human in thefuture.
So that's good.
That's like a great, interestingand useful thing.
However, the one thing that Ithink is missing is a lot of
(06:42):
things that humans don't reallyshow up on a job description on
a list of tasks, they're notalways even apparent.
Like I've had the experience andI'm sure most of our listeners
have had this experience, I'msure you have had where someone
leaves and you don't realizewhat they were doing until
(07:05):
they're gone.
One recent company I worked ateliminated about 35% of the
folk, the people there.
And there was a lot of chatteramong those of us who were
remaining where we would say toeach other did they know what
everyone was?
They know what some of thesepeople were doing.
And over the subsequent weeks,we found out, no, the company
was shocked to find out eachtime something would happen.
(07:30):
It was a task done or somethingdone by someone who's no longer
there?
Why is that happening?
Because it's very abstract.
It's very intangible.
A customer likes an employee alot, who's the customer service
person doesn't really show upanywhere.
A customer hears, Andreas orTodd on a podcast or a
prospective customer.
Here's Andreas, or Todd on apodcast and becomes the customer
(07:52):
that happened to be at Eightfoldsomeone from the military, the
Army, heard me on a webinar,contacted me, I got'em in touch
with sales folks.
And eightfold ended up gettingthis big thing called a Gig
Eagle, this big contract forArmy reservists technology that
helps find work for Armyreservists.
These are things that are likenot always on a list of a job
description or a list of tasks,and they're just these kind of
(08:15):
intangible things that happenover time where someone takes
these.
It takes on things that arehuman to human relationships
that are very hard to quantify.
I would think that's a veryimportant thing to consider as
you're breaking down jobs andpeople into tasks and roles
using AI.
Andreas Welsch (08:29):
I think that's a really important reminder,
right?
In this climate when you seelayoffs, restructuring left and
right in many differentindustries, certainly tech is
leading from what I'm seeing.
You do get the feeling that it'snumbers on a balance sheet in
some cases, right?
But there's a lot more, to yourpoint, than the sum of tasks or
(08:50):
line items in a job description.
Bob and Jane and Sarah do a lotmore than these five or 10
bullets here.
Todd Raphael (08:56):
Think about someone getting their hair done.
Could a robot eventually getsomeone's hair cut?
I'm sure.
Is that really why someone getstheir hair done?
Probably not.
They do it to talk to gossip, topour out their feelings,
whatever.
Some senior citizens get theirhair done every week or every
three weeks or whatever to getout of a senior citizen home and
get, take a little field trip,get just, it's something they
(09:16):
look forward to.
In that example, it's very, it'sunlikely that's on the job
description of the person.
Yes, there are many other jobsthat.
The human contact is much lessimportant.
Andreas Welsch (09:30):
So a couple of weeks I was on a podcast hosted
by a successful CEO of midsizebusiness, and we were talking
about agents as well.
And I said we're looking at thisright now in business a lot of
times as a cost exercise, can Itake out cost, move out people.
We no longer need them.
We have agents that can do thework, but what I feel we're
(09:50):
missing a lot of times isactually this enablement, this
multiplication type aspect.
What if we had the same amountof people, but with the help of
technology, they became moreproductive or more efficient, or
they had 10 times betterinsights and could help us grow
the business faster.
Yeah.
Because they know our industry,they know our customers, they
know our products.
(10:11):
Why is nobody thinking aboutthat?
And the answer that I got reallysurprised me because he said, if
I don't need more people, whyshould I hire them?
I don't hire more than I need.
And I said that's not the point,right?
The point is, you have people inyour business, you've already
invested in them, so why notgive them better tools and
better technology to help thembe more successful and help your
company be more successful?
That still baffles me that a lotof times this is the prevailing
(10:33):
mindset.
Todd Raphael (10:34):
Yeah.
Said.
Andreas Welsch (10:36):
Yeah.
Now, we've already talked abouta few gaps, I would say in this
vision of agents that we seeright now.
Another one that I'm seeing is,hey, there's not really an org
chart, right?
If we do compare agents andhumans or digital workers and
human workers in our humanworld, we have an art chart.
We know who reports to whom.
(10:57):
We know what the roles, what thejob functions are, we know what
the job descriptions are.
And yes, there are even moretasks between the lines that
employees do, but for agents intheory, anyone can spin up their
agent or their team of agent.
So what do you feel the orgchart looks like or is going to
look like as organizations bringmore agents into their business?
Todd Raphael (11:20):
Short answer.
I don't know.
As Peter Capelli, USC professorI talked to recently, reminded
me every prediction aboutbusiness is almost always wrong
because people just can't, youjust don't dunno what's gonna
happen in the future.
He gave an example of whenpeople predicted that truck
drivers would be obsolete andwe've already passed the date.
But if I had to predict, which Iwill, when we think about an org
(11:40):
chart, we think about most of uswould visualize boxes and lines,
and it's like a pyramid.
It's like a hierarchy.
And I don't know if thathierarchy really is gonna make
sense anymore because what we'rehearing or what's been talked
about is that in this new AIworld people's knowledge, their
(12:00):
experience, their insight, theirconnections, those four things I
think I listed were.
Those are gonna be superimportant.
In the current or old world itwas how many people do you
manage?
If you manage some people, youcan get a raise and then you can
manage some people with a highertitle.
Then you can maybe manage somemore people with a lofty title.
Then you get another raise orsomething like that's, and
(12:22):
somehow you're more important'cause you've managed people,
but.
What I'm hearing and reading isyour importance is partly gonna
be based on your knowledge, yourexperience, your network, your
connections, your understandingof an industry and the company.
That kind of stuff is notnecessarily about how many
people you manage when with whattitles, which tells me that
(12:43):
maybe the hierarchical boxeswith lines org chart maybe will
not be as valuable.
Maybe it'll be more like asports team or something where
like.
The manager, the coach, and thedugout, the sidelines of the, in
case of basketball or football,is not necessarily the most high
paid, or not necessarily themost famous or the most
valuable.
It's the individualcontributors, the players with
(13:05):
their skills and their knowledgeor whatever.
In the case of sports, I guessit would be skills, but the case
of corporate life, it might beyour knowledge and your
connections that might make youmore valuable than say, a
manager perhaps.
Andreas Welsch (13:16):
So a couple different thoughts there in,
things running through my mindas I'm listening to you say
this, and one is I wonder how doyou acquire that knowledge when.
(13:46):
People, when you as anindividual maybe do a bit more
of that cognitive offloading,relying on AI more and more, and
maybe a little too much, how doyou gain experience?
How do you become moreexperienced in your role and
eventually even if it's aflatter hierarchy, how do you
have opportunities and getopportunities to advance when
you work with AI as a co thinkeror as a, co-pilot in, in, in
(14:09):
that sense?
That's one.
A couple more, but
Todd Raphael (14:15):
I imagine it might depend on the industry.
If you were a, an expert inmedicine or restaurants or
retail or whatever it might be,and you came from some
background, whether as a nurseor a physician or a consultant
or a retail store manager andyou took on a job at some
company.
Your background might vary, butthe key here is that you have a
(14:37):
network, you have connections,and you have knowledge, you have
expertise.
You understand the field, thechallenges of it, and that kind
of stuff.
It's not so much that you arecoming to the company to manage
a certain number of people witha certain title or anything like
that.
It's what's in your head that'sharder to offload to a computer.
Andreas Welsch (14:57):
So basically what I think I'm hearing,
despite all of thattechnological support at our
fingertips and at ouravailability, prioritize the
human connection, prioritize thereal world experiences because
you can take those from task totask, from job to job in
eventually from company tocompany too, right?
That's what makes you valuable.
(15:17):
Your knowledge, your uniquecombination of skills and
network.
Todd Raphael (15:22):
I said again.
Yeah.
Andreas Welsch (15:23):
So one thing I'm wondering too, right?
I used to be a manager.
At one point I, I was managingmanagers who are managing teams.
When I read about flatterhierarchies, I put myself back
into those shoes from a coupleyears ago and I'd be pretty
afraid Hey, I've worked so hardto, claw my way up the org chart
and the hierarchy to get to thisposition, mid-level, senior
(15:45):
level leadership, and now you'retelling me the company no longer
needs me because there are fewerpeople that manage more agents.
Where is that going?
Especially when we talk aboutorg chart and flatter
hierarchies is that the realitythat you see that leaders need
to warm up to or get morecomfortable with?
Todd Raphael (16:06):
Yes.
I feel that when I see I see alot of people out there who are.
Valuable because of theirconnections, they're valuable
because on social media or onthe media or otherwise, they can
have an influence.
I see a lot of individualcontributors with a lot of
powerful value that maybe isgoing getting hidden because
(16:30):
they're not managing 14 people.
They don't have a lofty title orwhatever, but it's those super
connectors or those superknowledge folks, or those kind
of people who are hiddenconnectors, hidden sources of
knowledge hidden sources ofinstitutional knowledge of an
organization who are so oftenunderused, underpaid, and
(16:53):
underappreciated.
Andreas Welsch (16:55):
So it sounds like we need to shift again to
impact as opposed to just beingthere and doing the work that
you're assigned, but what is theimpact you're actually having?
Who are you influencing?
What are you influencing?
More on, on that side whenagents can take over some of the
tasks that we've been doing.
Todd Raphael (17:11):
And so often it's off the radar screen.
I got a note.
I'd say it was a few months agofrom someone at Eightfold asking
me a bunch of questions aboutsome work that I was doing there
that they were now assigned to.
I haven't been there in since2022.
I left to go to Sky Hive.
Been there in three years and noone knew that I was doing this
(17:34):
work that I had taken on in myown interest.
And, a lot, and that's a verytangible example, but there's a
lot of relationships and otherkinds of things that just fly
under the radar that are veryhard to quantify and put your
finger on till someone leaves.
Yeah.
Andreas Welsch (17:49):
Yeah.
Reminds me of some of the thingsthat, that I remember when I was
in, in corporate, things thatI've done or that I had done 10
years ago or put in place or inpractice were still being used.
By the time I was leaving withlittle updates or a little
maintenance even though, otherpeople were taking it over so
you it all the time.
It's a good point.
Yeah.
Todd Raphael (18:08):
Just a reminder of this story and if you've been to
like a lot of human resourcesand talent and workforce
conferences.
You've heard this story and it'sprobably getting really old, and
if you haven't been, it'sprobably interesting.
But so I've heard this story afew times, but Disney, the
company did this study wherethey were trying to figure out
they decided they could only putso much eggs in one basket as
(18:30):
far as hiring and managingpeople.
They can't, like not everysingle person in the company is.
The most important person in thecountry company by definition.
So they try, they set out to tryto figure out, what are the most
important jobs in this companythat we really need to focus on?
Where a slightly better employeeis gonna be so much different
than a slightly worse employeeas opposed to just, it doesn't
(18:51):
make a ton of difference ifsomeone's a little better, worse
at something.
It is it the animators?
Is it the people whose checksafety on amusement park rides?
Is it an actor?
Although I guess that was a badexample.
'cause the safety checkers areprobably important.
Who can deny that?
But anyhow, they came out withthe people at Disneyland or
Disney World who are dressed as.
(19:13):
Mickey Mouse or Goofy orsomething in the, and the reason
they came out with that beingthe most important job is that
those people are the ones whoare most often asked, where's
the bathroom?
Where's the shortest line?
Where's the, when's the sungonna go down on this ride?
It's too hot.
Or, we're gonna get a waterfountain and like that.
(19:36):
Customer experience is likesuper hugely everything.
It's gonna make or breaksomeone's day, the bathroom, the
lines, the water, but yet thepeople's job is to dress up as
Mickey and run around, dowhatever you do as Mickey.
A really powerful example of howthese things are intangible and
offshoots of our regular dayjobs.
Andreas Welsch (19:54):
That's a great example.
So I guess I haven't been toenough HR conferences yet.
I didn't my assumption was, yeahit's the experience.
You wouldn't want to have Mickeyor Pluto or any of the other
characters be in a bad mood.
That's why it's important.
But certainly that the tangiblethings, where can I get water or
Yeah.
Things matter.
Absolutely.
Now, to me it feels like theindustry has to stick with the
(20:17):
ride in an amusement parkexample, right?
Shot through the rollercoasterof the AI hype and the agent AI
hype.
And we've quickly acceleratedto, to get to the peak where I
think we're right now, justbefore the peak, over the last
12 months or so.
But I also see a lot of theconversation is actually driven
by te.
You have vendors that whisper inyour CIO's ear, Hey, come build
(20:39):
here, build on my platform.
Or we've built something intoour applications that you can
use.
But a lot of times who I'mmissing and who I feel is not
part of those conversations,actually HR leaders, right?
So I think as an HR leader, youdon't just need to be in the
building.
You need to actually have a seatat the table when these
conversations are taking place.
(21:01):
How do you recommend HR leadersget involved when it's a lot of
times the IT or technologydepartment leading these agenda
AI initiatives?
What do HR leaders need to do?
Todd Raphael (21:12):
For one thing in the future, these may become one
and the same as has been talkedabout.
You're managing.
Work.
You're not managing people,you're not managing computers,
you're managing work.
And so a, it might be that insome companies IT and HR kind of
become highly intertwined.
You're managing work, whether aperson does it or whether a
(21:34):
computer does it, or whether acombination of a person or a
computer does.
So I guess one answer to yourquestion would be there may not
be this distinction at somepoint in the future, But
ultimately I think that humanresources people may be able to
see the big picture better thansome other people.
'cause they might be able to seeon the entire lifecycle of a
person.
You know what I'm saying?
If we use.
Ai and it's part of the process.
(21:55):
Let's say that we are usingartificial intelligence to hire
people better based on theirskills, not based on where they
went to college or whe orwhether they even went to
college.
Let's say that they're gonna tryto do that and use AI to find
skilled people who maybe didn'tgo to college or did but right.
But make sure they particularlydon't ignore those who didn't go
to college.
So if that's.
What is going on?
(22:16):
The HR folks might be the oneswho best understand how to look
at the entire public, entireemployee life cycle and say,
okay, what kind of skills do weneed?
Not just for hiring, but oncethese people are on the staff,
how might people be upskilled orreskilled?
How do we keep these people fromleaving?
How do we, design anorganization that make people
(22:37):
really engaged and motivated?
'cause it's a two-way street.
You can find the best employeeever and they just are gonna
lose interest if you're, if it'sa, poor environment.
So there's all kinds of aspectsto the workplace that from soup
to nuts, from start to finish,from identifying a candidate to
the time they're.
Interviewed, assessed, selected,hired, motivated, stay managed,
(23:00):
skilled, re-skilled, moved intoother jobs, take on gigs, taking
on mentors in-house learningfrom a coach in-house.
There's so many parts of thatperhaps human resources folks
are well situated to see thatbig picture.
And they also, a bit of along-winded answer, but they
(23:20):
also are probably well situatedto have data that show some of
these things.
So they might be able to showhow recruiting connects to
retention.
They might be able to go, Hey,look, our data show that if you
hire this person that you'resaying is great, that person's
likely to leave in two and ahalf months.
So let's hire someone who wouldbe a great employee, meaning
(23:40):
they're going to be great andwant to stay.
So they might have that kind ofdata that someone else might not
have at their disposal.
'cause again, they might have alot of workforce related
insights from their technology.
Andreas Welsch (23:53):
Fantastic.
That ties it together so nicely,especially mentioning all the
different functions andprocesses that HR supports.
All of them being critical fromhire to retire and talent
development and successionplanning and everything.
Wonderful.
Todd I've really been lookingforward to our conversation.
I've been looking for someone tohave a conversation like this
(24:13):
with in the HR space for quitesome time.
And it's been a real challengefinding people who are
knowledgeable, who have aperspective on this.
So I'm super excited that we gotto chat today.
Me too.
Before we close it, I waswondering if you can summarize
the key three takeaways for ouraudience from today.
We covered a lot of ground overthe last 25 minutes.
What would you say are the keythree things that our audience
(24:36):
members should know?
Todd Raphael (24:37):
Learn from learn more about your people and what
they truly, what truly they're,what these true human qualities
that they have are not vagarieslike empathy, but what truly are
they doing for customers or forbringing in customers or other
things could be related totechnology.
What is it they're doing that'smaybe not apparent or not on
(24:58):
the.
On the stat sheet, on their jobdescription or something like,
really learn deeply about yourpeople best.
You can use technology to helpplan for the future to see what
can computers do in the future.
What can people do in thefuture?
How might we plan our workforcearound that?
Some jobs may be combined withother jobs.
Some jobs can be partiallyautomated.
How will that change the jobonce it's partially automated?
(25:19):
So that is number two, to use,your workforce.
Planning to think about therelationship between how AI is
gonna affect computers andpeople.
In the future.
And then I'm thinking on thespot here, but so bear with me.
But number three is just thetwo-way street really.
That the, or really the employeelifecycle being not just
(25:43):
something that you can't take asliver and say, we're gonna now
hire people based on their,based on skills, not based on
their pedigree.
You need to look at the entireemployee life cycle.
Hiring, learning, retention,turnover, re-skilling, mentors,
internal talent marketplaces.
So people taking on gigsinternally.
(26:04):
Look at that entire marketplaceand think about all of how AI
can holistically change andimprove everything you do from
soup to nuts with withemployees.
Andreas Welsch (26:13):
Beautiful.
Thank you so much, Todd.
Thanks for joining and forsharing your expertise with us.
Like I said, it was a realpleasure having you on and
learning from you, especially asthis space is so nascent and
emerging yet moving so quickly.
So really appreciate having youon today.
Todd Raphael (26:28):
Thanks.
Hey, it's the end of 2025 and things are
getting real.
What actually changes when youbring agents into a business?
How do teams need to be set upand what do HR and business
leaders need to get we'll talkabout all that.
So let's jump right in.
Welcome to What's the BUZZ?,where leaders share how they
have turned hype into outcome.
(00:52):
Today, we'll talk about how todesign your workforce for
agentic AI and who better totalk about it than someone who's
actively working in this space.
Todd, Raphael.
Todd, thank you so much forjoining.
Todd Raphael (01:02):
Thanks for having me.
Andreas Welsch (01:03):
Hey, Todd, maybe not everybody in the audience is
familiar with you or has comeacross your content.
I certainly have, and I've beena big fan of the things that you
put out for a long time.
Why don't you tell us a littlebit about yourself, who you are
and what you do?
Todd Raphael (01:15):
No problem.
I spent more than a coupledecades at a couple
organizations Cranes and thenERE on conferences and
publications for recruiters, HRprofessionals and others.
And then went to a HR technologycompany called Eightfold fairly
early on and helped grow thatmarketing department, grow the
organization dramatically overthree years.
(01:37):
And then was recruited to go toa competing organization called
Sky Hive, which sold toCornerstone.
Cornerstone on demand.
About a couple years ago, andI've been working as a helping
HR tech companies, talentacquisition technology companies
and related companies, staffingcompanies and so on over the
last couple years, help figureout how to differentiate
(02:00):
themselves, how to communicatewhat they do, how to explain
what they do communicate, createcontent, work with influencers
and that kind of thing.
Help them figure themselves outand how to find themselves and
communicate who they are.
Andreas Welsch (02:12):
Thank you for sharing.
I feel a lot of times marketinggets a bit of a bad reputation
and I've previously been in, inmarketing myself is in my
career.
But I feel you also, get a verygood grasp on what's actually
happening in, in the market.
What are different vendorsdoing, what do buyers actually
look for, which is why I'm soexcited to have you on the show
and learn from your perspectiveas, as well what you're seeing.
(02:35):
Todd, it's the end of 2025.
It feels like we've talked aboutAI agents as an industry for, I
don't know a year or so.
It feels like it's already beena decade in AI years.
And it feels like now we'rereally at this point where we're
seeing more and moreorganizations, more and more
leaders, more and more HRleaders also come to the table
(02:55):
and say what do we do?
We're spinning up individualproofs of concepts and pilots
and we're working with differentvendors to try things out, and I
think that's great.
Finally we're moving to towardsaction, but when I talk to
leaders, when I attend vendorconferences and it's probably
similar for you as well, itseems that everybody just talks
(03:17):
about the building part ofagents, but there's not a lot of
talk about what actually happensafterward.
What happens when you bring themto the organization?
What happens when they are inproduction?
I'm curious with all this talkabout did your human employees
get agent employees?
Do we now need to draw parallelsto how we manage humans as well?
Can we apply some of that toagents?
(03:38):
What are you seeing?
Todd Raphael (03:39):
Yeah, I've been to a bunch of conferences this last
couple years and including tworecently.
And there was all kinds ofbooths having to do with agents
and buying agents.
And one of'em, you get to buildyour own agents right at the
booth there on a little iPad,and then after five minutes
you're done, you got your agent.
But yeah, there was not a wholelot of explanation exactly.
What does your day look like?
What is it?
How is this agent kind of workin your workflow?
(04:02):
Like you get up, you have a job,you find out there's someone
quit, so you have an open job.
How do you fill that job usingan agent?
What do you do?
Like step A, step B, step C,step D and E.
Like, where's the agent fittingin and what do they do and what
do they not do?
And then once that agent's partof your workflow, your hiring
process.
(04:22):
What then do you do?
How do you know in six months ifit's biased or how do you know
if it's accurate?
It may be perfectly fair, butnot very accurate.
You may be not really an imp,maybe not be an improvement over
how things were without an agentor maybe there's some
personalization with the agentthat's no longer working very
effectively or something.
But, so those are, I think, openquestions, but certainly there's
(04:43):
not a lot of informationexactly.
About how do you spend a day orspend a week or a month.
With agents, what exactly, howdoes that work in your life,
your work life, your job?
Andreas Welsch (04:52):
Yeah.
I mean I've been creating somecourses with LinkedIn Learning
on the topic what happens whenyou bring agents to business.
And some of them came out at theend of last year.
So even very early in, in theirwave.
But I'm sure things have beenevolving and especially as you
and your businesses are thinkingabout this too.
My, my recommendation would be,think about it.
How, yeah.
(05:13):
How do we use these things once,once they're in our business in
that sense.
And I feel with all that focuson building, to your point,
there's not enough conversationabout what happens afterwards.
Yeah.
And that, that opens, manyparallels in my point of view to
how do we work with people,right?
We wanna make sure thatdepending on the role or the
(05:33):
company that we run, backgroundchecks are you who you say you
are.
Permissions for your role.
You get certain permissions tocertain systems, but not more
extensive than what you reallyneed to guidelines.
Codes of conduct.
I want you to behave in acertain way.
I'm thinking about what iftechnology teams build agents in
every different department orevery user builds it slightly
(05:54):
differently.
Do we need to ground them insome sort of code of conduct,
for example?
What do you think?
Todd Raphael (06:00):
I think that there's this idea that, so a lot
of companies I work with areworking on talent intelligence
or skills intelligence, andthey're taking AI and they're
taking jobs and trying to figureout what tasks and skills are
involved in each job.
That by breaking a job up intotasks and skills, you can help
(06:21):
predict the future.
You can say, oh, this job, halfof it may no longer be necessary
to be done by a human, or thesetwo jobs could be melded
together in the human in thefuture.
So that's good.
That's like a great, interestingand useful thing.
However, the one thing that Ithink is missing is a lot of
(06:42):
things that humans don't reallyshow up on a job description on
a list of tasks, they're notalways even apparent.
Like I've had the experience andI'm sure most of our listeners
have had this experience, I'msure you have had where someone
leaves and you don't realizewhat they were doing until
(07:05):
they're gone.
One recent company I worked ateliminated about 35% of the
folk, the people there.
And there was a lot of chatteramong those of us who were
remaining where we would say toeach other did they know what
everyone was?
They know what some of thesepeople were doing.
And over the subsequent weeks,we found out, no, the company
was shocked to find out eachtime something would happen.
(07:30):
It was a task done or somethingdone by someone who's no longer
there?
Why is that happening?
Because it's very abstract.
It's very intangible.
A customer likes an employee alot, who's the customer service
person doesn't really show upanywhere.
A customer hears, Andreas orTodd on a podcast or a
prospective customer.
Here's Andreas, or Todd on apodcast and becomes the customer
(07:52):
that happened to be at Eightfoldsomeone from the military, the
Army, heard me on a webinar,contacted me, I got'em in touch
with sales folks.
And eightfold ended up gettingthis big thing called a Gig
Eagle, this big contract forArmy reservists technology that
helps find work for Armyreservists.
These are things that are likenot always on a list of a job
description or a list of tasks,and they're just these kind of
(08:15):
intangible things that happenover time where someone takes
these.
It takes on things that arehuman to human relationships
that are very hard to quantify.
I would think that's a veryimportant thing to consider as
you're breaking down jobs andpeople into tasks and roles
using AI.
Andreas Welsch (08:29):
I think that's a really important reminder,
right?
In this climate when you seelayoffs, restructuring left and
right in many differentindustries, certainly tech is
leading from what I'm seeing.
You do get the feeling that it'snumbers on a balance sheet in
some cases, right?
But there's a lot more, to yourpoint, than the sum of tasks or
(08:50):
line items in a job description.
Bob and Jane and Sarah do a lotmore than these five or 10
bullets here.
Todd Raphael (08:56):
Think about someone getting their hair done.
Could a robot eventually getsomeone's hair cut?
I'm sure.
Is that really why someone getstheir hair done?
Probably not.
They do it to talk to gossip, topour out their feelings,
whatever.
Some senior citizens get theirhair done every week or every
three weeks or whatever to getout of a senior citizen home and
get, take a little field trip,get just, it's something they
(09:16):
look forward to.
In that example, it's very, it'sunlikely that's on the job
description of the person.
Yes, there are many other jobsthat.
The human contact is much lessimportant.
Andreas Welsch (09:30):
So a couple of weeks I was on a podcast hosted
by a successful CEO of midsizebusiness, and we were talking
about agents as well.
And I said we're looking at thisright now in business a lot of
times as a cost exercise, can Itake out cost, move out people.
We no longer need them.
We have agents that can do thework, but what I feel we're
(09:50):
missing a lot of times isactually this enablement, this
multiplication type aspect.
What if we had the same amountof people, but with the help of
technology, they became moreproductive or more efficient, or
they had 10 times betterinsights and could help us grow
the business faster.
Yeah.
Because they know our industry,they know our customers, they
know our products.
(10:11):
Why is nobody thinking aboutthat?
And the answer that I got reallysurprised me because he said, if
I don't need more people, whyshould I hire them?
I don't hire more than I need.
And I said that's not the point,right?
The point is, you have people inyour business, you've already
invested in them, so why notgive them better tools and
better technology to help thembe more successful and help your
company be more successful?
That still baffles me that a lotof times this is the prevailing
(10:33):
mindset.
Todd Raphael (10:34):
Yeah.
Said.
Andreas Welsch (10:36):
Yeah.
Now, we've already talked abouta few gaps, I would say in this
vision of agents that we seeright now.
Another one that I'm seeing is,hey, there's not really an org
chart, right?
If we do compare agents andhumans or digital workers and
human workers in our humanworld, we have an art chart.
We know who reports to whom.
(10:57):
We know what the roles, what thejob functions are, we know what
the job descriptions are.
And yes, there are even moretasks between the lines that
employees do, but for agents intheory, anyone can spin up their
agent or their team of agent.
So what do you feel the orgchart looks like or is going to
look like as organizations bringmore agents into their business?
Todd Raphael (11:20):
Short answer.
I don't know.
As Peter Capelli, USC professorI talked to recently, reminded
me every prediction aboutbusiness is almost always wrong
because people just can't, youjust don't dunno what's gonna
happen in the future.
He gave an example of whenpeople predicted that truck
drivers would be obsolete andwe've already passed the date.
But if I had to predict, which Iwill, when we think about an org
(11:40):
chart, we think about most of uswould visualize boxes and lines,
and it's like a pyramid.
It's like a hierarchy.
And I don't know if thathierarchy really is gonna make
sense anymore because what we'rehearing or what's been talked
about is that in this new AIworld people's knowledge, their
(12:00):
experience, their insight, theirconnections, those four things I
think I listed were.
Those are gonna be superimportant.
In the current or old world itwas how many people do you
manage?
If you manage some people, youcan get a raise and then you can
manage some people with a highertitle.
Then you can maybe manage somemore people with a lofty title.
Then you get another raise orsomething like that's, and
(12:22):
somehow you're more important'cause you've managed people,
but.
What I'm hearing and reading isyour importance is partly gonna
be based on your knowledge, yourexperience, your network, your
connections, your understandingof an industry and the company.
That kind of stuff is notnecessarily about how many
people you manage when with whattitles, which tells me that
(12:43):
maybe the hierarchical boxeswith lines org chart maybe will
not be as valuable.
Maybe it'll be more like asports team or something where
like.
The manager, the coach, and thedugout, the sidelines of the, in
case of basketball or football,is not necessarily the most high
paid, or not necessarily themost famous or the most
valuable.
It's the individualcontributors, the players with
(13:05):
their skills and their knowledgeor whatever.
In the case of sports, I guessit would be skills, but the case
of corporate life, it might beyour knowledge and your
connections that might make youmore valuable than say, a
manager perhaps.
Andreas Welsch (13:16):
So a couple different thoughts there in,
things running through my mindas I'm listening to you say
this, and one is I wonder how doyou acquire that knowledge when.
(13:46):
People, when you as anindividual maybe do a bit more
of that cognitive offloading,relying on AI more and more, and
maybe a little too much, how doyou gain experience?
How do you become moreexperienced in your role and
eventually even if it's aflatter hierarchy, how do you
have opportunities and getopportunities to advance when
you work with AI as a co thinkeror as a, co-pilot in, in, in
(14:09):
that sense?
That's one.
A couple more, but
Todd Raphael (14:15):
I imagine it might depend on the industry.
If you were a, an expert inmedicine or restaurants or
retail or whatever it might be,and you came from some
background, whether as a nurseor a physician or a consultant
or a retail store manager andyou took on a job at some
company.
Your background might vary, butthe key here is that you have a
(14:37):
network, you have connections,and you have knowledge, you have
expertise.
You understand the field, thechallenges of it, and that kind
of stuff.
It's not so much that you arecoming to the company to manage
a certain number of people witha certain title or anything like
that.
It's what's in your head that'sharder to offload to a computer.
Andreas Welsch (14:57):
So basically what I think I'm hearing,
despite all of thattechnological support at our
fingertips and at ouravailability, prioritize the
human connection, prioritize thereal world experiences because
you can take those from task totask, from job to job in
eventually from company tocompany too, right?
That's what makes you valuable.
(15:17):
Your knowledge, your uniquecombination of skills and
network.
Todd Raphael (15:22):
I said again.
Yeah.
Andreas Welsch (15:23):
So one thing I'm wondering too, right?
I used to be a manager.
At one point I, I was managingmanagers who are managing teams.
When I read about flatterhierarchies, I put myself back
into those shoes from a coupleyears ago and I'd be pretty
afraid Hey, I've worked so hardto, claw my way up the org chart
and the hierarchy to get to thisposition, mid-level, senior
(15:45):
level leadership, and now you'retelling me the company no longer
needs me because there are fewerpeople that manage more agents.
Where is that going?
Especially when we talk aboutorg chart and flatter
hierarchies is that the realitythat you see that leaders need
to warm up to or get morecomfortable with?
Todd Raphael (16:06):
Yes.
I feel that when I see I see alot of people out there who are.
Valuable because of theirconnections, they're valuable
because on social media or onthe media or otherwise, they can
have an influence.
I see a lot of individualcontributors with a lot of
powerful value that maybe isgoing getting hidden because
(16:30):
they're not managing 14 people.
They don't have a lofty title orwhatever, but it's those super
connectors or those superknowledge folks, or those kind
of people who are hiddenconnectors, hidden sources of
knowledge hidden sources ofinstitutional knowledge of an
organization who are so oftenunderused, underpaid, and
(16:53):
underappreciated.
Andreas Welsch (16:55):
So it sounds like we need to shift again to
impact as opposed to just beingthere and doing the work that
you're assigned, but what is theimpact you're actually having?
Who are you influencing?
What are you influencing?
More on, on that side whenagents can take over some of the
tasks that we've been doing.
Todd Raphael (17:11):
And so often it's off the radar screen.
I got a note.
I'd say it was a few months agofrom someone at Eightfold asking
me a bunch of questions aboutsome work that I was doing there
that they were now assigned to.
I haven't been there in since2022.
I left to go to Sky Hive.
Been there in three years and noone knew that I was doing this
(17:34):
work that I had taken on in myown interest.
And, a lot, and that's a verytangible example, but there's a
lot of relationships and otherkinds of things that just fly
under the radar that are veryhard to quantify and put your
finger on till someone leaves.
Yeah.
Andreas Welsch (17:49):
Yeah.
Reminds me of some of the thingsthat, that I remember when I was
in, in corporate, things thatI've done or that I had done 10
years ago or put in place or inpractice were still being used.
By the time I was leaving withlittle updates or a little
maintenance even though, otherpeople were taking it over so
you it all the time.
It's a good point.
Yeah.
Todd Raphael (18:08):
Just a reminder of this story and if you've been to
like a lot of human resourcesand talent and workforce
conferences.
You've heard this story and it'sprobably getting really old, and
if you haven't been, it'sprobably interesting.
But so I've heard this story afew times, but Disney, the
company did this study wherethey were trying to figure out
they decided they could only putso much eggs in one basket as
(18:30):
far as hiring and managingpeople.
They can't, like not everysingle person in the company is.
The most important person in thecountry company by definition.
So they try, they set out to tryto figure out, what are the most
important jobs in this companythat we really need to focus on?
Where a slightly better employeeis gonna be so much different
than a slightly worse employeeas opposed to just, it doesn't
(18:51):
make a ton of difference ifsomeone's a little better, worse
at something.
It is it the animators?
Is it the people whose checksafety on amusement park rides?
Is it an actor?
Although I guess that was a badexample.
'cause the safety checkers areprobably important.
Who can deny that?
But anyhow, they came out withthe people at Disneyland or
Disney World who are dressed as.
(19:13):
Mickey Mouse or Goofy orsomething in the, and the reason
they came out with that beingthe most important job is that
those people are the ones whoare most often asked, where's
the bathroom?
Where's the shortest line?
Where's the, when's the sungonna go down on this ride?
It's too hot.
Or, we're gonna get a waterfountain and like that.
(19:36):
Customer experience is likesuper hugely everything.
It's gonna make or breaksomeone's day, the bathroom, the
lines, the water, but yet thepeople's job is to dress up as
Mickey and run around, dowhatever you do as Mickey.
A really powerful example of howthese things are intangible and
offshoots of our regular dayjobs.
Andreas Welsch (19:54):
That's a great example.
So I guess I haven't been toenough HR conferences yet.
I didn't my assumption was, yeahit's the experience.
You wouldn't want to have Mickeyor Pluto or any of the other
characters be in a bad mood.
That's why it's important.
But certainly that the tangiblethings, where can I get water or
Yeah.
Things matter.
Absolutely.
Now, to me it feels like theindustry has to stick with the
(20:17):
ride in an amusement parkexample, right?
Shot through the rollercoasterof the AI hype and the agent AI
hype.
And we've quickly acceleratedto, to get to the peak where I
think we're right now, justbefore the peak, over the last
12 months or so.
But I also see a lot of theconversation is actually driven
by te.
You have vendors that whisper inyour CIO's ear, Hey, come build
(20:39):
here, build on my platform.
Or we've built something intoour applications that you can
use.
But a lot of times who I'mmissing and who I feel is not
part of those conversations,actually HR leaders, right?
So I think as an HR leader, youdon't just need to be in the
building.
You need to actually have a seatat the table when these
conversations are taking place.
(21:01):
How do you recommend HR leadersget involved when it's a lot of
times the IT or technologydepartment leading these agenda
AI initiatives?
What do HR leaders need to do?
Todd Raphael (21:12):
For one thing in the future, these may become one
and the same as has been talkedabout.
You're managing.
Work.
You're not managing people,you're not managing computers,
you're managing work.
And so a, it might be that insome companies IT and HR kind of
become highly intertwined.
You're managing work, whether aperson does it or whether a
(21:34):
computer does it, or whether acombination of a person or a
computer does.
So I guess one answer to yourquestion would be there may not
be this distinction at somepoint in the future, But
ultimately I think that humanresources people may be able to
see the big picture better thansome other people.
'cause they might be able to seeon the entire lifecycle of a
person.
You know what I'm saying?
If we use.
Ai and it's part of the process.
(21:55):
Let's say that we are usingartificial intelligence to hire
people better based on theirskills, not based on where they
went to college or whe orwhether they even went to
college.
Let's say that they're gonna tryto do that and use AI to find
skilled people who maybe didn'tgo to college or did but right.
But make sure they particularlydon't ignore those who didn't go
to college.
So if that's.
What is going on?
(22:16):
The HR folks might be the oneswho best understand how to look
at the entire public, entireemployee life cycle and say,
okay, what kind of skills do weneed?
Not just for hiring, but oncethese people are on the staff,
how might people be upskilled orreskilled?
How do we keep these people fromleaving?
How do we, design anorganization that make people
(22:37):
really engaged and motivated?
'cause it's a two-way street.
You can find the best employeeever and they just are gonna
lose interest if you're, if it'sa, poor environment.
So there's all kinds of aspectsto the workplace that from soup
to nuts, from start to finish,from identifying a candidate to
the time they're.
Interviewed, assessed, selected,hired, motivated, stay managed,
(23:00):
skilled, re-skilled, moved intoother jobs, take on gigs, taking
on mentors in-house learningfrom a coach in-house.
There's so many parts of thatperhaps human resources folks
are well situated to see thatbig picture.
And they also, a bit of along-winded answer, but they
(23:20):
also are probably well situatedto have data that show some of
these things.
So they might be able to showhow recruiting connects to
retention.
They might be able to go, Hey,look, our data show that if you
hire this person that you'resaying is great, that person's
likely to leave in two and ahalf months.
So let's hire someone who wouldbe a great employee, meaning
(23:40):
they're going to be great andwant to stay.
So they might have that kind ofdata that someone else might not
have at their disposal.
'cause again, they might have alot of workforce related
insights from their technology.
Andreas Welsch (23:53):
Fantastic.
That ties it together so nicely,especially mentioning all the
different functions andprocesses that HR supports.
All of them being critical fromhire to retire and talent
development and successionplanning and everything.
Wonderful.
Todd I've really been lookingforward to our conversation.
I've been looking for someone tohave a conversation like this
(24:13):
with in the HR space for quitesome time.
And it's been a real challengefinding people who are
knowledgeable, who have aperspective on this.
So I'm super excited that we gotto chat today.
Me too.
Before we close it, I waswondering if you can summarize
the key three takeaways for ouraudience from today.
We covered a lot of ground overthe last 25 minutes.
What would you say are the keythree things that our audience
(24:36):
members should know?
Todd Raphael (24:37):
Learn from learn more about your people and what
they truly, what truly they're,what these true human qualities
that they have are not vagarieslike empathy, but what truly are
they doing for customers or forbringing in customers or other
things could be related totechnology.
What is it they're doing that'smaybe not apparent or not on
(24:58):
the.
On the stat sheet, on their jobdescription or something like,
really learn deeply about yourpeople best.
You can use technology to helpplan for the future to see what
can computers do in the future.
What can people do in thefuture?
How might we plan our workforcearound that?
Some jobs may be combined withother jobs.
Some jobs can be partiallyautomated.
How will that change the jobonce it's partially automated?
(25:19):
So that is number two, to use,your workforce.
Planning to think about therelationship between how AI is
gonna affect computers andpeople.
In the future.
And then I'm thinking on thespot here, but so bear with me.
But number three is just thetwo-way street really.
That the, or really the employeelifecycle being not just
(25:43):
something that you can't take asliver and say, we're gonna now
hire people based on their,based on skills, not based on
their pedigree.
You need to look at the entireemployee life cycle.
Hiring, learning, retention,turnover, re-skilling, mentors,
internal talent marketplaces.
So people taking on gigsinternally.
(26:04):
Look at that entire marketplaceand think about all of how AI
can holistically change andimprove everything you do from
soup to nuts with withemployees.
Andreas Welsch (26:13):
Beautiful.
Thank you so much, Todd.
Thanks for joining and forsharing your expertise with us.
Like I said, it was a realpleasure having you on and
learning from you, especially asthis space is so nascent and
emerging yet moving so quickly.
So really appreciate having youon today.
Todd Raphael (26:28):
Thanks.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Are You A Charlotte?
In 1997, actress Kristin Davis’ life was forever changed when she took on the role of Charlotte York in Sex and the City. As we watched Carrie, Samantha, Miranda and Charlotte navigate relationships in NYC, the show helped push once unacceptable conversation topics out of the shadows and altered the narrative around women and sex. We all saw ourselves in them as they searched for fulfillment in life, sex and friendships. Now, Kristin Davis wants to connect with you, the fans, and share untold stories and all the behind the scenes. Together, with Kristin and special guests, what will begin with Sex and the City will evolve into talks about themes that are still so relevant today. "Are you a Charlotte?" is much more than just rewatching this beloved show, it brings the past and the present together as we talk with heart, humor and of course some optimism.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.