May 23, 2025 • 37 mins
Patty Hayward is the General Manager of Healthcare and Life Sciences at Talkdesk, where she leads innovation at the intersection of AI, technology, and patient experience for healthcare organizations.
Key Takeaways
- Deep Personal Motivation: Patty’s passion for patient experience stems from her family’s struggle navigating complex healthcare for her mother. This drives her mission to simplify and humanize patient journeys.
- Transforming Contact Centers: Talkdesk is applying AI to rethink traditional healthcare contact centers. By integrating electronic health records, smart routing, and agent assist tools, they aim to ensure every patient feels known, supported, and guided from their first call.
- Empowering Agents: The goal is to upskill contact center agents-reducing time-to-proficiency, increasing job satisfaction, and creating healthcare career growth. AI handles routine queries, while agents focus on complex, human-centered needs.
- AI for Patient Experience: Patty draws compelling analogies to Amazon and Delta, envisioning healthcare systems that anticipate patient needs and minimize frustration. She argues that a truly great patient experience requires patients never have to repeat themselves.
- Critical AI Adoption Considerations: The democratization of AI (especially LLMs) is making advanced solutions accessible to organizations without deep IT resources. Yet, Patty stresses the importance of rethinking processes-“don’t automate a bad process”-and keeping trust, privacy, and clinical care at the forefront.
- Balancing Innovation and Trust: Patty acknowledges that while rapid AI adoption is exciting, healthcare’s primary goal is always quality clinical care. Any patient engagement technology must build trust, ensure safety, and ultimately improve-not complicate-outcomes.
About Patty Hayward
Patty Hayward, GM of healthcare and life sciences at Talkdesk, has over a quarter century of industry strategy experience, including at organizations such as McKesson, Medicity, and Humedica. She is an expert in HIE, population health, pharmacy, process redesign for healthcare systems, and increasing access to patient information.
A "glow up" signifies a positive transformation, reflecting the journey of becoming a better, more successful version of oneself.
At The Tech Glow Up, we humanize the startup and innovation landscape by focusing on the essential aspects of the entrepreneurial journey. Groundbreaking ideas are often ahead of their time, making resilience and perseverance vital for founders and product leaders.
In our podcast, we engage with innovators to discuss their transformative ideas, the challenges they face, and how they create value for future success.
If you're a founder or product leader seeking your own glow up, or a seasoned entrepreneur with stories to share, we invite you to join our guest list via this link.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Nathan C (00:00):
Hello and welcome to The Glow Up Fabulous
conversations with innovativeminds.
Today I'm talking with PattyHayward, who is the general
manager of healthcare and lifesciences at Talkdesk.
Patty, it's so good to see youtoday.
Thanks for joining.
Thank you for having me!
Patty Hayward (00:16):
It's great to be here.
Nathan C (00:17):
As somebody who is just fascinated with the story
of innovation and getting thingsdone at enterprise scale, maybe
even in healthcare, I am soexcited to hear about your
journey! Can you introduceyourself and tell us a little
bit about what you do atTalkdesk?
Patty Hayward (00:36):
Yeah! Absolutely.
So I've been at Talkdesk forabout four and a half years now,
and I started when, that companyreally wanted to specifically
look at some verticals.
Healthcare and life science is areally different space and sort
of your traditional horizontaltake on technology.
They really wanted to understandhow to be successful there.
So I came in and helped guidedthe strategy, you know, what
(01:00):
should we build?
How should we build it?
What are we going to do as faras go-to-market messaging?
What's working today inhealthcare?
What's broken?
How do we go about reallygetting people to transform how
they think about patientexperience?
As you think about all thetechnology that's coming out,
there's a lot around theclinical side of things.
We really focus on the patientexperience side.
Nathan C (01:21):
I'm eager to know because customer and patient
experience is like one of thethings that gets me excited most
about technology.
What got you interested inadvocating for and applying
technology to the patientexperience?
As you said, there's so manyproblems to solve in this space.
Patty Hayward (01:39):
I've been in healthcare for 25 years, so I've
done a lot of different things,from pharmacy to looking at
things like value-based care, alot of different things of
intersection of value and whatwe can do to help patients
better serve themselves, havebetter outcomes, things like
that.
And that really all started withan experience I had with my mom
(02:01):
who had multiple chronicdiseases and navigating
healthcare is really, reallyhard.
And My perspective was no oneshould have to go through what
we went through as far as how toget her access to care.
How to ensure that theirphysicians were coordinating
well.
How do we even ensure that shedoesn't wind up in the ICU for
(02:21):
five days as what happened atone point.
And that disease eventually tookher life.
So you know, to me it was,there's so much that we can do
that's better.
There's a lot of things wecouldn't change, because you
know, these chronic diseases aredegenerative in nature, but we
certainly don't need to beadditive in some of the things
that we do, and the complexityis just off the hook.
(02:42):
So working between both theprovider side working with our
payer organizations as well asour big pharma and distribution
organizations and really helpingto understand and really to
bring to light some newopportunities to do better, has
really been great because reallyunderstanding across the
spectrum, again, the complexityis huge and trying to understand
(03:05):
what's at the root cause of it,and then what can we change?
Because a lot of this isn't justabout technology.
It's about the people, theprocess, and the technology
triangle that we have.
Nathan C (03:15):
I'm always reminded when I ask this question that
there's almost always a personalstory, and a very personal pain
that drives people to work on aproblem so expansive and, over
such a long period of time.
I'm a little stunned to hear asenior leader in healthcare
talking about how horrible astruggle it was for them to work
(03:40):
through that process and some ofthe parts of the healthcare
system that we hear so muchabout.
Right?
You know, chronic care,value-based care.
so thank you for that.
As you're looking at things likenew technologies, I'm sure
artificial intelligence, largelanguage models and others have
got to be in that toolkit oftechnologies that you're using.
(04:05):
Can you talk a little bit aboutthe core problem that you're
working on now at Talkdeskaround these patient
experiences, and what sorts oftechnologies are you bringing in
to really make a differencethere?
Patty Hayward (04:17):
So if you think about, and the nice thing about
doing what I do is everybody canrelate because we've all called
a physician's office, we'veprobably all called our health
insurance or tried to dosomething on their website,
right?
That experience, especially whenI started, it's getting better,
but it's pretty universally bad.
And we all have highexpectations because we're kind
(04:38):
of spoiled by Amazon and Uberand some of these other
applications, that have reallyallowed us to have a much easier
experience doing things that wetend to do every day.
Calling a cab, buying anythingreally on Amazon.
And so those are the experiencesthat we'd like to see across
other sectors.
(04:58):
And that was kind of how I cameinto this was, when someone
approached me about contactcenter, I kind of was cocked my
head and went,"I'm not sure whyI would be interested in working
around contact center." But Iwill tell you that I started to
really think about the patientexperience and I started to say,
"You know what, there's nowhereto go but up!" So how do we go
and impact that in a meaningfulway?
(05:20):
And you know, I always come withthe personal perspective of,
I'll call my primary carephysician, which I've been with
for 20 years now, and every timeI call, it's like the first time
I've called, right?
I'm brand new every time becausethey don't know who I am.
They don't know my journey, andI have to say all the things all
over again, every time.
So that was the first thing Iwanted to solve was: How do we
(05:41):
connect the system of record tothe agents that are answering
the phone so that they're nothaving to go and look up things
and doing all sorts of thingsand asking you all those
questions that are quite franklyannoying.
How can you go in and get thatinformation by linking into the
electronic health record,bringing that in, making sure
we're doing more than just ascreen pop.
Doing things like when you callDelta Airlines and you are gonna
(06:05):
fly them that day and yourflight is delayed, they know who
you are.
They know the fact that it'sdelayed, and that's probably why
you're calling.
So how do we take those in thosebits of information and turn
them into an experience thatfeels like, okay, we know you.
Now let's route you.
We know what you're callingabout.
We know the problem that we'retrying to solve.
If we can't solve it with, youknow, an agentic agent, right?
(06:29):
So someone that's artificialintelligently driven, which
they're getting better andbetter and better, who do we get
them to that can answer theirquestion the first time so that
we're not transferring to 10different people?
We're not further sort ofantagonizing you as you're
calling in to try to do things.
And typically when people arecalling in healthcare, they're
(06:49):
calling for some reason thatusually increases their stress
level.
So, you know, the stress piecethat all kind of comes to light
a lot when you're calling andtrying to deal with healthcare.
So it's really important to makepatients feel really known and
supported and help guide them.
My goal is I want them to beable to be a care guide, right?
(07:10):
You're probably having toescalate to a human because you
have a complex problem and youneed guidance.
And so we want to empower thoseagents with information at the
touch, at the ready, as they'relistening to the conversations
to provide them answers with nothaving them,"Hold please.
Let me go look up the answer.""Hold please.
Let me go talk to so-and-so andget back." Hopefully all of this
(07:32):
information is presented in realtime.
It's recorded for qualitypurposes, and we can help train
other people as they come in.
You know, I've worked withorganizations that tell me it
takes them three months beforethey allow an agent to answer a
call from a patient.
Yet their average tenure ofthose agents is 18 months.
(07:52):
That is a really tough situationfor an organization to deal
with.
So how do we help them A.) Getagents up to speed B.) Can we
empower those agents to do moreso that they have more
satisfaction in their job andthey're not just an ask- answer
machine?
But they're really trying tosolve problems so that we help
them, you know, we say work atthe top of the license for
pharmacists, nurses, doctors.
(08:14):
We should do the same thing foragents.
Let's train them! Let's havedifferent ladders of what they
can accomplish and give them acareer path that feels worthy.
You know, people go intohealthcare because they wanna
serve, and so being able toallow them to do that in
meaningful ways that get betterand better and more complex, as
all of this AI gets better andcan solve the basic problems,
(08:36):
that's what I think the goal is.
I don't think we'll ever get tothe point where we don't have
people answering phone calls inhealthcare.
It's too personal and there'stoo many complicated things that
we need to deal with.
But I think being able to havethose folks be much more
educated, much more empowered todo things and really help move
(08:57):
patients through doing what theyneed to do, to do the best
things for their health, that'sa win-win for everybody.
Nathan C (09:04):
I love how we started with the patient pain- point and
addressed how agents and a humanagent response could really
start to move those patientsinto, better response times,
better information, lessfrustration.
(09:25):
But then it very quickly alsobecame about upskilling.
About addressing, you know,that's a pretty quick turn on a
labor force, especially ifyou're spending a third of the
time on training and they're notactually doing any calls.
Very interesting to see how,ideally, improvements in that
patient care chain also startsto improve the experience for
(09:47):
the workforce.
This idea that people could beworking in their strengths and
helping, right?
Rather than kind of pushing itdown the road, or reading policy
descriptions about why thingsare unknown.
It feels way more satisfying,fulfilling of a role there.
Patty Hayward (10:04):
And think about
Nathan C (10:05):
Amazing!
Patty Hayward (10:05):
how we look up information now.
Like it used to be Google whereyou had tons of links and you
just had to sort of sort throughand read everything and try to,
you know, put it together inyour own head.
But if you look at ChatGPT andhow it delivers information for
you in a much moreconversational, like, here's a
summary, you know, and you canask it questions and it'll
continue to give you more andmore information.
(10:26):
That's similar to what you'reseeing now for agents being
guided.
I will tell you when I started,I sat down with agents'cause you
know, you wanna understand whattheir a day in the life looks
like for them.
And I remember sticky notes allover the screens, dual screens
with like Google sheets behindwith all sorts of information
because of the complexity ofscheduling a physician
(10:47):
appointment.
It's not like going and gettingyour hair done.
People are like,"Why can't weautomate more patient
scheduling?" That's a, a verylayered discussion.
There's lots of reasons,probably some that really aren't
valid and need to be changed.
And that is our biggest blockerto being able to really adopt
(11:07):
more digital technologies, etcetera.
You know, if you look at otherindustries, when was the last
time you talked to a human in abank?
And that's a highly regulatedindustry, right?
So we're a highly regulatedindustry as well.
There's a lot we can learnthere.
It's not the same.
There's a lot of things that aredifferent.
One of the main differences thatI always like to remind my
leadership team of is (11:26):
Why aren't people adopting more
rapidly in healthcare?
The central tenant forhealthcare is good clinical
care.
It's not customer service.
It's clinical care.
We want to make sure that thatpatient is receiving the best
clinical care possible, and thatis first and foremost our
guiding light.
(11:47):
If we can also deliver greatpatient experience, wonderful,
we can do that.
But first and foremost, it'sgreat clinical care.
So it has taken time to get to apoint where this technology has
advanced enough that they canput the limited resources that
they have to do the ancillarygoals.
And we've sort of democratizedAI now with a lot of these large
(12:08):
language models because they canbuild themselves.
That is transformational forthese organizations who don't
have the IT resources to throwat, you know, coding and complex
AI.
You know, scientific work thatneeds to be done to train them.
So it's a lot easier now tostart to consume this.
(12:30):
Now, we have a long way to gowith trust and you know, how do
we start thinking about biasesand privacy and all of these
different aspects that we haveto be good stewards of inside of
healthcare.
And that's the nice thing aboutpatient experiences because it's
not clinical care, right?
You're not giving clinicaladvice.
(12:51):
And so that is a little easierplace to start than something
like diagnoses.
And those are moving reallyrapidly too.
So as you think through all ofthese different areas, you can't
walk down a trade show nowwithout seeing AI or agentic AI
all over the place.
Hims was, every single booth hadsomething about AI and agents
(13:13):
and all of those things.
So, separating out what'smarketing speak and what's real
is also a difficult thing for alot of these organizations and
they need a lot of help.
Nathan C (13:22):
One of the things that I noticed when I was at
HLTH in October, people are veryexcited to use technology to get
the busy work out of the way ofcare.
There's so much in the way ofthe workforce, let's just make
it easy for everybody please.
But this idea of how do youtrust?
(13:42):
How do you build?
How do you use AI and agentictools in the right places that
foster trust, that giveappropriate information?
I always talk about inmarketing, automation
accelerates everything in yourbusiness, especially the
dysfunction.
AI is really good atacceleration and does things
(14:04):
that takes us months in blinks.
How do you balance, innovation,the development costs like you
were noting.
Plus, very important thingsabout trust, safety, patient
care and data?
There's so many things that youare touching and impacting.
Patty Hayward (14:21):
Yeah, for sure! And you know, I think it's
important that we understandtechnology can't solve
everything.
When you automate a bad process,you just go faster with the bad
process.
So it's very important when westart a project with our
customers to take a step backand say,"Don't just pour old
wine into a new bottle." Youneed to rethink what you're
doing and how you're doing it sothat you can actually hit the
(14:44):
goals that we've set out,because any organization that's
buying something wants to lookat a return on investment,
right?
I mean, you don't spend moneywithout understanding how you're
gonna get your money back.
And so being able to hit thosereturn on investment metrics and
goals is really important.
And in order to do that, youreally need to look at your
processes, your culture, howyou're going about embracing the
(15:06):
technology, because technologycan't fix your core
dysfunctions.
You've gotta identify them andlook at ways that you can change
your processes and your culturein order to optimize the
technology.
I used to sell robots back inthe day that would do
centralized cart fill, and so itwas a really dumb machine.
(15:28):
It would just pick things up andit scanned the barcode.
It knew what it was picking up,and it would pick it and bring
it over and drop it into anenvelope.
If you don't optimize yourprocesses around what it does
well, you're not gonna do wellwith it.
Like if you wanna make it dosomething that it's not
programmed to do or in a waythat it's not optimized, you're
not gonna have the return.
And it was pretty black andwhite.
(15:49):
We had organizations that weretransformational with it.
Got all of their pharmacists outon the floors, you know, and
really were able to change theirentire culture.
Others wanted it to just doexactly what it's always done
and they didn't get the returns.
So it's really, really justimportant to do that.
Now, AI is much smarter so youcan bend it a little bit more to
your will, but then you end upwith still a lot of issues with
(16:14):
things like, okay, I have a lotof physicians that want Gladys
to still answer their phone'cause Gladys has always
answered their phone and sheknows my routine for scheduling
and how I like to be scheduledand all of this and that.
And when everything is aone-off, nothing can be
automated.
We need strong physicianleadership, right?
We need strong support of theautomation to say, these are the
(16:37):
things guys, if we can all cometogether and figure out a way
where we can sort of optimizehow we schedule.
We've had patient portalscheduling for decades! And when
was the last time you were ableto schedule online with a
doctor?
Hardly ever.
There's a few that have adoptedit, but very, very rarely.
So it's not a technology issue,it's a how do we get folks to
(16:59):
really think differently aboutthe way that they're doing
things?
And there's some greatorganizations, I've been really
leaning in to organizations thatare seeing success and
understanding, okay, how are youguys doing these things?
And really leaning into thephysician leadership, because
that's really important tounderstand a, you know, what do
they need in order to practicewell, we wanna support them.
(17:20):
That's, that's a crucial thing.
Remember, central tenantclinical care, right?
So how do we support them?
But how do we also take cost outof the business?
We all are feeling the absoluteincrease in costs around
healthcare.
It's mind-boggling how muchhealthcare costs now versus 10,
20 years ago.
So we need to do some things torein it in.
And that might mean we need tomake some changes to the type of
(17:43):
appointments that you have anddo that in a way that allows you
to still have and honor yourspecialty practice.
But at the same time allowsautomation to occur so that the
organization can reduce theiroverhead.
It's the only way we're gonnastop this train wreck of
healthcare cost increases.
So how do we continue to do thatin a way that's collaborative,
(18:05):
transformational, and reallyembracing the new technology
that the organizations havedecided to purchase.
Nathan C (18:11):
Goodness! This is like such a masterclass in
change and transformation.
Really doesn't matter how muchtechnology or how powerful of a
technology you're layering in,right?
If it's not Supporting the coregoals of the user, if it doesn't
(18:33):
have buy-in at an incrediblyhigh level early on.
When you're looking at measuringthe impact of the work that you
do today is cost really theprimary metric or, it sounds
like you're working, a lotdirectly with doctors
potentially within systems andlarger organizations.
(18:55):
How do you help them?
I think as somebody who's intransformation, this is
important, right?
Like how do you help the peoplewho are undergoing
transformation measure theimpact of it?
And then how do you measure yourimpact within that process and
that growth.
Patty Hayward (19:11):
So I think it's twofold, right?
So we do have, of course,standard like time savings
measurements, things like that.
We call it full containmentinside of a contact center,
which means we've been able tofully answer a question from,
you know, an automatedperspective, we're not, having
to transfer to a human.
So there's a big cost savingsthere, obviously.
Speed to answer.
So you've reduced the queue sizebecause you're taking off the
(19:34):
easy questions, and now you'renot having to wait a half an
hour to talk to a human being,right?
So reduction in wait time.
Big patient satisfier, obviouslythat's a huge thing.
A lot of times you're in reallycompetitive spaces where you've
got five or six health systemsthat are competing for the
business of that patient.
And if they can't get through orcan't get an actual appointment,
(19:55):
they're gonna go someplace elsewhere they can.
So, you know, whoever has thefirst opening gets the business
type of thing.
So it's important to make surethat we can address those needs,
making sure that we can helpoptimize, how fast we can get to
that patient, work with them onscheduling, making it easy for
them to schedule and reschedule.
That's really important.
And then in the member side, ifyou think about the payer side,
(20:18):
especially for MedicareAdvantage, which you do a lot of
work with those organizations,they get ranked on star ratings
and member satisfaction is ahuge piece of that.
So being able to ensure thatthey're being answered right
away, they're being thoroughlyanswered in the language that
they can speak.
All of those different thingsare really, really important to
(20:39):
these payer organizations.
And then when you think aboutlife sciences and you think
about especially a lot of thescrutiny that they get under for
pricing and different thingslike that, how do we help their
patient access centers empowerthose patients that need to take
the drugs that they've broughtto market but can't afford.
And that's their role is to takethose referrals from the
(20:59):
physicians who want to help themand get them on therapy.
How do we keep them on therapy?
Because we know it can be lifechanging, but we also know some
of these specialty drugs are$10,000 a dose.
How can we help them afford thatand truly understand what
they're supposed to do'causesome of these drugs, if you
don't take them exactly asprescribed when they are
prescribed, they're ineffectiveand they're also$10,000 a dose.
(21:21):
So how do you go about makingsure that these folks are
really, really in tune with whatthey're supposed to do when
they're supposed to do it, andthat they're helped with the
cost?
So across those different worldsare sort of what we do and can
intersect and help, to reallyallow these organizations to do
that and look across at patientsatisfaction, healthcare
(21:41):
provider satisfaction as thesehealthcare providers are calling
in member satisfaction.
Those are high ranking needs forthese guys as well as of course,
all the efficiency metrics.
How can we make sure wait timesare lower?
How can we make sure that you'reanswering on the first call as
opposed to having them call backmultiple times'cause they get
frustrated without getting theanswer that they think they
need.
(22:02):
You know, we call that"firstcall resolution." So there's a
lot of standard metrics, KPIsthat we have in contact center.
But I've always challenged myorganizations to think a little
differently too, around how dowe think about number of
appointments scheduled?
Are we able to look at, hey, youknow, as you're starting to
create time and space for theseagents, are we answering more
(22:24):
than just the question theycalled in for?
An example would be,"I've runout of refills on my
prescription." and The agent cansay,"Okay, I can help you with
that.""You haven't been in foryour annual physical.
You need to get your mammogram,you need to get your A1C
checked." This list of thingsthat are also happens to be the
value-based care metrics thatthe health system is measured
(22:45):
upon, and that's how they gettheir bonus payments.
Or conversely, they get knockedfor not meeting those metrics
that they need.
And they end up gettingpenalized.
So how do we take them frompenalty to bonus?
So that helps their bottom lineand it becomes a value driver
for everyone because if they dothe things that we've identified
in those HEDIS measures, we knowthat they're gonna stay
(23:07):
healthier.
They're gonna stay outta thehospital.
They're not gonna have theacuity that you would have if
you just ignore.
We get a lot of peopleunfortunately, who don't do the
things that they're supposed todo, and they don't come in on a
regular basis either from fear,from just cost all these things.
How can we understand what'sgoing on here and how can we
help them better afford a drugthat maybe they're only, they're
(23:29):
supposed to be taking once aday?
Now they're only taking it everyother day.
Is that cost?
Is that side effects?
How do we know that?
That's a big thing.
Patient days on therapy is ameasure for these organizations.
So how do we make sure thatthey're doing that and that we
understand the reasons whythey're not, if they're not.
Are they having problems gettingto the pharmacy?
Should we just mail it to them?
What are the ways that we canhelp these patients do the
(23:52):
things that they need to do tostay healthier, which reduces
cost, which helps with bonuspayments, which helps with a
patient experience.
Talk about the ultimate patientexperience if you stay healthy.
Lack of health is the worstdriver of mental health
problems, quality of life, etcetera.
I think being able to help guidethese organizations to guide
their patients throughcomplicated therapies and
(24:15):
protocols and appointmentscheduling and all those things,
that to me is my ultimate visionof where we go when we get sort
of time back for these agents.
Nathan C (24:24):
I'm bought on a hundred percent.
You know this, this picture thatyou painted of from the first
time I call into a call centerto schedule an appointment or to
get a clarification on somethingI'm now kind of in the care of
this agentic support, which hasa little bit of visibility into
(24:47):
the history, can start to bethat through line that used to
be the very friendly woman whoanswers the phone, for all the
doctors and really like startsto hold onto things, be a
consistent data set through thatwhole experience.
And I love this idea that, notjust providing value at that
first engagement, but deliveringengaging, teaching back into.
(25:11):
Anything that gives doctors moretime to doctor, I think the
whole system would be thrilledwith! Absolutely! So I love this
idea of learning and growing,you kinda have to paint the
picture for a team and encouragethem to go find the path, even
when it's not all there today.
So, the show is called The GlowUp.
(25:32):
A"glow up" is a notabletransformation, or a rebirth.
We use it to talk about bigaudacious goals that you might
have in the next six months,When you're thinking about
transformation in healthcare, Ifeel like any six month goal is
audacious on its own! What sortof audacious goals do you have
for your work at Talkdesk inhealth and life sciences?
Patty Hayward (25:53):
So I'll take it into two sectors because I think
there's the technology goals,which are fun and really
inspiring right now about whatwe can do.
And then there's thetransformation, sort of like,
okay, what do we need to do forthem to be able to take
advantage in a meaningful way ofthis technology?
As we've talked about, you'vegotta have both.
I think from a technologyperspective, I think this
(26:13):
agentic AI and the things thatwe are building and doing.
It's just like nothing I've everseen as far as how fast these
things are evolving and theabilities that they have with
sort of multiple, each agentkind of has its own specialty,
just like you do with humans.
And then you've got thissupervisor agent that can then
say,"Okay, I see you're callingfor this.
(26:35):
I'm gonna send you to thisspecialty agent who's now gonna
ask you the questions." You'renot gonna see that it's not
gonna transfer, it's just gonnaautomatically do that so that it
knows which group, which thingsto ask.
And that way you're notconfusing things, you're not
getting that hallucination.
All those different things thatwe've been able to solve, and
allow those to be deployed bothin a digital way, right?
(26:55):
So typing in any channel youwant, right?
So not everybody wants to call.
I know my daughter would love tonever speak to another human
again.
She'd love to text everybody.
So then there's the, you know,my dad who wants to call
everybody and I'm probablysomewhere in the middle.
And I think that, being able tomeet people in their channel of
(27:15):
choice, have those conversationsin a meaningful way in the
language that they speaknatively without having to have
them translate.
That might cause its own set ofproblems.
So being able to translate intothe language that they are
native to and being able to,really help folks.
That for me, from atechnological standpoint, the
next six months we're gonna seesome just amazing output there.
(27:37):
So I'm really excited for that.
And then on the transformationside, I think I talked a lot
about sort of what I'm reallyfocused on today, which is
really working withorganizations to look at what
are the blockers, how do we needto transform the culture and the
processes that they've builtover all of these years that may
not serve well with automation,right?
(28:01):
As people are starting to buymore and more into like, look,
we've gotta transform.
These are our importantfinancial goals that we have to
achieve and it's important forus to ladder up and down from
C-Suite all the way through tothe end users and the
physicians, et cetera.
Everybody's gotta kind of berowing in the same boat and
that's not easy! Buy in and ofitself, but then to kind of see
(28:23):
how do we make these changes?
What are some of the things thatwe can experiment with and try
maybe in small pods that aremore open to change.
What are the ones that we canuse?
Let's try.
Fail fast because these kinds ofmodels allow us to fail fast.
That didn't work?
Okay, let's try this tweak.
Ooh, that worked! And you can doa lot more changes more easily
(28:43):
with this work than in the pastwhere it required so much coding
and resources, you don't needthat as much anymore.
So it's much easier for us to doiterations and to iterate fast
to get to that ultimate, okay,this is working.
This is working really well.
So that's kind of thetransformational work I'm really
excited about in the next sixmonths, that I'm working on now.
Nathan C (29:02):
Do you use agents in that like alignment and
discovery and process work thatyou do on the transformation
side as well?
Patty Hayward (29:15):
When you say agents, are you talking about
human agents or are you talkingabout the agentic AI?
Nathan C (29:19):
Does a agentic AI support those initiatives?
Patty Hayward (29:22):
So yes, the agentic AI, so there's the
agents that answer the phone,and then there's the ones behind
the scenes that sort of analyzepatterns and different things
that they're seeing and hearing.
I have an example of one of ourfinancial services group.
They did some amazing work andused our platform that listens
and finds patterns indiscussions and things like that
(29:44):
to detect fraud in real time.
They noticed a certain patternwhen someone was calling and
trying to take over someone'saccount, and they were able to
put together a real time alertto their supervisors when they
detected a pattern like this,that they would come in and cut
it off in real time, and theywere able to completely reduce
(30:06):
the amount of money that wasbeing stolen out of different
bank accounts, by the way thatthese guys were trying to reset
passwords, right?
So I'm trying to reset apassword.
They call in with things likecrying baby in the background,
right?
Like,"I've gotta get this doneright away.
I'm late for work, my baby'scrying." You make this really
stressful situation for theagents, and the agents kind of
comply and do it.
(30:27):
They started to see that patternand they were able to cut this
off in real time.
So it's pretty exciting to seesome of the evolutions of how
this can do it.
We know in healthcare we'reunder attack on a regular basis
from people trying to doransomware for that data.
So password reset's a big dealright now.
How do we detect and look atfraudulent ways of this stuff
(30:49):
happening that we can cut off inreal time.
How can we do things like makesure that when we do have an
agent that's having some troublecommunicating in real time or
alert a supervisor to be able tokind of come in and join and
support that.
Lots of great things that we cando in real time now, that I
think are gonna get better andbetter and better.
You talked about AI editing.
It can detect and really look atthings in really good ways.
(31:12):
You can tell it,"Eliminate allthe ahs and ums and awkward
silences and things like that."That's the way things are
working in detecting.
Because we record all of thesecalls for quality purposes,
right?
How do we go about learning andtraining our agents to do
better?
Learning and training ouragentic agents to be better?
What are areas folks are gettingstuck and they have to escalate?
(31:34):
Maybe we don't need that tohappen.
So how do we iterate?
Iteration and quick iteration isreally important in this
innovation.
Nathan C (31:41):
Oh my goodness! So many seeds there! What if I'm
having a frustratedconversation, right, sort of
unpacking your fraud example,and the agentic AI could help
the agent understand thisparticular patient is getting
very frustrated.
Could you send them some flowersor whatever that digital
(32:01):
equivalent is, or maybe justtake a break and take a breath.
The ability for AI to sort ofhelp those of us who are
managing so much detail tounderstand what's important in
the moment, and to see the otherinputs that might be driving the
problem is such a powerfulthing, to think about.
(32:23):
I
Patty Hayward (32:23):
love it because it automatically takes notes and
summarizes things for me.
I'm a talk thinker.
use my hands a lot.
Trying to write things down mynotes are terrible.
They're universally terrible.
I have like one word that I waslike, for some reason I thought
that was important at one point,but I don't actually know what
it means anymore.
My notes tend to be a lot likethat because I get really
involved in a conversation andI'm really trying to actively
listen as opposed to justwriting all the stuff down.
(32:46):
This allows us to not have to dothat, right?
It can guide more and it cangive you less programmatic
responses to things which Ithink
Nathan C (32:53):
Cool.
is really going
Patty Hayward (32:54):
to help all of that.
Nathan C (32:55):
So on each episode of The Glow Up, we do make an
opportunity for a communityshout out or spotlight to a
group that's doing work for goodor impact.
Is there a spotlight that you'dlike to share your episode
today?
Patty Hayward (33:08):
Thanks for that opportunity.
April's Parkinson's AwarenessMonth and my sister-in-law
suffers from this awful disease.
She's quite young.
She's in her early fifties, soit's pretty devastating.
The support and the researchthat's going on right now, talk
about transformation, there's alot of great things that are
going on in scientific research,so I always like to highlight
the Parkinson's ResearchFoundation.
(33:30):
I think they are incredible.
They're doing great things.
They're funding a lot of amazingresearch that I would love to
see continue.
Nathan C (33:36):
Heck yeah! Thank you so much and so timely.
Sort of switching gears a littlebit, in your work at Talkdesk
and the things that you'relooking to accomplish over the
next year or so, is thereanything that you're looking
for, whether it's partners,research, team members, et
cetera?
Patty Hayward (33:54):
I think I'm gonna go back to that partnering with
the physician organizations.
I would love to get a focusgroup around just that.
And bring some really keyphysician leaders together to
really think through how do wethink about automation in a
different way?
And how can we makerecommendations and have
multiple types ofrecommendations depending on the
(34:15):
organization that would allowboth their practice to thrive as
well as automation to continueto evolve so that there's more
and more things that we can do,overheads reduce all of those
different great things.
I'm very clear on clinical careis the central tenet.
That's really important.
We don't wanna change that.
We're not looking to take thataway.
(34:35):
And then how do we go forwardlooking from that?
So that's been my primary focusfor the last several months as
I've started to have a lot offolks go live with the new
technology and sort of start torun into some blockers of
adoption.
Primary care is pretty simple todo'cause it's a pretty standard
type of care.
But once you get into theologies and all of the different
(34:57):
types of specialties out there,and then you have subspecialties
and subspecialties, so muchcomplexity! Don't even get me
started on radiology.
You know, how do we go aboutscheduling all of these things
in a way that allows us to honorthe needs of making sure folks
have prior authorizations andall of those different things,
but at the same time, allows formore automation so that we can
(35:18):
do more and more things like wetalked about with care guiding.
Nathan C (35:21):
I love that the call to action is around building
partnerships and doing consumerresearch.
Could not be more on brand forThe Glow Up! I love it! Patty,
it has been such a delight tochat with you.
Thank you for spending so muchof your time today, with us on
The Glow Up.
If people wanted to learn moreor join in a research
(35:42):
partnership, what's the best wayto follow up with you or connect
online?
Patty Hayward (35:47):
Probably LinkedIn.
That is probably the best placeto go.
I publish all of my thoughtleadership and podcasts and
different things that we'redoing at Talkdesk there, so it's
a great place to see what'sgoing on and to connect with me.
Nathan C (35:59):
Patty Hayward, it has been such a fantastic treat for
me to talk about AI drivenhealthcare transformation with
you.
I really appreciate so much yourfocus on That central goal of
working in the healthcare space,which is provide top quality
(36:19):
care and to power, peopleproviding it to do the best,
they can with so manyconstraints and so many
challenges.
I'm really heartened to hear theAI story in healthcare moving
beyond just like,"Hey, we'll fixyour data." but We'll actually
like activate your data, empoweryou, and really start to build
(36:43):
this model for a unifiedcontinuum of care, not just like
40 different disparateconversations where you say 60%
of the same thing over and over.
I've never been so excited abouthealthcare scheduling and
supporting doctors withautomation.
Thank you so much for thefantastic work that you're doing
(37:05):
and sharing it with us on TheGlow Up.
Patty Hayward (37:07):
Absolutely! Thanks for having me!
Hello and welcome to The Glow Up Fabulous
conversations with innovativeminds.
Today I'm talking with PattyHayward, who is the general
manager of healthcare and lifesciences at Talkdesk.
Patty, it's so good to see youtoday.
Thanks for joining.
Thank you for having me!
Patty Hayward (00:16):
It's great to be here.
Nathan C (00:17):
As somebody who is just fascinated with the story
of innovation and getting thingsdone at enterprise scale, maybe
even in healthcare, I am soexcited to hear about your
journey! Can you introduceyourself and tell us a little
bit about what you do atTalkdesk?
Patty Hayward (00:36):
Yeah! Absolutely.
So I've been at Talkdesk forabout four and a half years now,
and I started when, that companyreally wanted to specifically
look at some verticals.
Healthcare and life science is areally different space and sort
of your traditional horizontaltake on technology.
They really wanted to understandhow to be successful there.
So I came in and helped guidedthe strategy, you know, what
(01:00):
should we build?
How should we build it?
What are we going to do as faras go-to-market messaging?
What's working today inhealthcare?
What's broken?
How do we go about reallygetting people to transform how
they think about patientexperience?
As you think about all thetechnology that's coming out,
there's a lot around theclinical side of things.
We really focus on the patientexperience side.
Nathan C (01:21):
I'm eager to know because customer and patient
experience is like one of thethings that gets me excited most
about technology.
What got you interested inadvocating for and applying
technology to the patientexperience?
As you said, there's so manyproblems to solve in this space.
Patty Hayward (01:39):
I've been in healthcare for 25 years, so I've
done a lot of different things,from pharmacy to looking at
things like value-based care, alot of different things of
intersection of value and whatwe can do to help patients
better serve themselves, havebetter outcomes, things like
that.
And that really all started withan experience I had with my mom
(02:01):
who had multiple chronicdiseases and navigating
healthcare is really, reallyhard.
And My perspective was no oneshould have to go through what
we went through as far as how toget her access to care.
How to ensure that theirphysicians were coordinating
well.
How do we even ensure that shedoesn't wind up in the ICU for
(02:21):
five days as what happened atone point.
And that disease eventually tookher life.
So you know, to me it was,there's so much that we can do
that's better.
There's a lot of things wecouldn't change, because you
know, these chronic diseases aredegenerative in nature, but we
certainly don't need to beadditive in some of the things
that we do, and the complexityis just off the hook.
(02:42):
So working between both theprovider side working with our
payer organizations as well asour big pharma and distribution
organizations and really helpingto understand and really to
bring to light some newopportunities to do better, has
really been great because reallyunderstanding across the
spectrum, again, the complexityis huge and trying to understand
(03:05):
what's at the root cause of it,and then what can we change?
Because a lot of this isn't justabout technology.
It's about the people, theprocess, and the technology
triangle that we have.
Nathan C (03:15):
I'm always reminded when I ask this question that
there's almost always a personalstory, and a very personal pain
that drives people to work on aproblem so expansive and, over
such a long period of time.
I'm a little stunned to hear asenior leader in healthcare
talking about how horrible astruggle it was for them to work
(03:40):
through that process and some ofthe parts of the healthcare
system that we hear so muchabout.
Right?
You know, chronic care,value-based care.
so thank you for that.
As you're looking at things likenew technologies, I'm sure
artificial intelligence, largelanguage models and others have
got to be in that toolkit oftechnologies that you're using.
(04:05):
Can you talk a little bit aboutthe core problem that you're
working on now at Talkdeskaround these patient
experiences, and what sorts oftechnologies are you bringing in
to really make a differencethere?
Patty Hayward (04:17):
So if you think about, and the nice thing about
doing what I do is everybody canrelate because we've all called
a physician's office, we'veprobably all called our health
insurance or tried to dosomething on their website,
right?
That experience, especially whenI started, it's getting better,
but it's pretty universally bad.
And we all have highexpectations because we're kind
(04:38):
of spoiled by Amazon and Uberand some of these other
applications, that have reallyallowed us to have a much easier
experience doing things that wetend to do every day.
Calling a cab, buying anythingreally on Amazon.
And so those are the experiencesthat we'd like to see across
other sectors.
(04:58):
And that was kind of how I cameinto this was, when someone
approached me about contactcenter, I kind of was cocked my
head and went,"I'm not sure whyI would be interested in working
around contact center." But Iwill tell you that I started to
really think about the patientexperience and I started to say,
"You know what, there's nowhereto go but up!" So how do we go
and impact that in a meaningfulway?
(05:20):
And you know, I always come withthe personal perspective of,
I'll call my primary carephysician, which I've been with
for 20 years now, and every timeI call, it's like the first time
I've called, right?
I'm brand new every time becausethey don't know who I am.
They don't know my journey, andI have to say all the things all
over again, every time.
So that was the first thing Iwanted to solve was: How do we
(05:41):
connect the system of record tothe agents that are answering
the phone so that they're nothaving to go and look up things
and doing all sorts of thingsand asking you all those
questions that are quite franklyannoying.
How can you go in and get thatinformation by linking into the
electronic health record,bringing that in, making sure
we're doing more than just ascreen pop.
Doing things like when you callDelta Airlines and you are gonna
(06:05):
fly them that day and yourflight is delayed, they know who
you are.
They know the fact that it'sdelayed, and that's probably why
you're calling.
So how do we take those in thosebits of information and turn
them into an experience thatfeels like, okay, we know you.
Now let's route you.
We know what you're callingabout.
We know the problem that we'retrying to solve.
If we can't solve it with, youknow, an agentic agent, right?
(06:29):
So someone that's artificialintelligently driven, which
they're getting better andbetter and better, who do we get
them to that can answer theirquestion the first time so that
we're not transferring to 10different people?
We're not further sort ofantagonizing you as you're
calling in to try to do things.
And typically when people arecalling in healthcare, they're
(06:49):
calling for some reason thatusually increases their stress
level.
So, you know, the stress piecethat all kind of comes to light
a lot when you're calling andtrying to deal with healthcare.
So it's really important to makepatients feel really known and
supported and help guide them.
My goal is I want them to beable to be a care guide, right?
(07:10):
You're probably having toescalate to a human because you
have a complex problem and youneed guidance.
And so we want to empower thoseagents with information at the
touch, at the ready, as they'relistening to the conversations
to provide them answers with nothaving them,"Hold please.
Let me go look up the answer.""Hold please.
Let me go talk to so-and-so andget back." Hopefully all of this
(07:32):
information is presented in realtime.
It's recorded for qualitypurposes, and we can help train
other people as they come in.
You know, I've worked withorganizations that tell me it
takes them three months beforethey allow an agent to answer a
call from a patient.
Yet their average tenure ofthose agents is 18 months.
(07:52):
That is a really tough situationfor an organization to deal
with.
So how do we help them A.) Getagents up to speed B.) Can we
empower those agents to do moreso that they have more
satisfaction in their job andthey're not just an ask- answer
machine?
But they're really trying tosolve problems so that we help
them, you know, we say work atthe top of the license for
pharmacists, nurses, doctors.
(08:14):
We should do the same thing foragents.
Let's train them! Let's havedifferent ladders of what they
can accomplish and give them acareer path that feels worthy.
You know, people go intohealthcare because they wanna
serve, and so being able toallow them to do that in
meaningful ways that get betterand better and more complex, as
all of this AI gets better andcan solve the basic problems,
(08:36):
that's what I think the goal is.
I don't think we'll ever get tothe point where we don't have
people answering phone calls inhealthcare.
It's too personal and there'stoo many complicated things that
we need to deal with.
But I think being able to havethose folks be much more
educated, much more empowered todo things and really help move
(08:57):
patients through doing what theyneed to do, to do the best
things for their health, that'sa win-win for everybody.
Nathan C (09:04):
I love how we started with the patient pain- point and
addressed how agents and a humanagent response could really
start to move those patientsinto, better response times,
better information, lessfrustration.
(09:25):
But then it very quickly alsobecame about upskilling.
About addressing, you know,that's a pretty quick turn on a
labor force, especially ifyou're spending a third of the
time on training and they're notactually doing any calls.
Very interesting to see how,ideally, improvements in that
patient care chain also startsto improve the experience for
(09:47):
the workforce.
This idea that people could beworking in their strengths and
helping, right?
Rather than kind of pushing itdown the road, or reading policy
descriptions about why thingsare unknown.
It feels way more satisfying,fulfilling of a role there.
Patty Hayward (10:04):
And think about
Nathan C (10:05):
Amazing!
Patty Hayward (10:05):
how we look up information now.
Like it used to be Google whereyou had tons of links and you
just had to sort of sort throughand read everything and try to,
you know, put it together inyour own head.
But if you look at ChatGPT andhow it delivers information for
you in a much moreconversational, like, here's a
summary, you know, and you canask it questions and it'll
continue to give you more andmore information.
(10:26):
That's similar to what you'reseeing now for agents being
guided.
I will tell you when I started,I sat down with agents'cause you
know, you wanna understand whattheir a day in the life looks
like for them.
And I remember sticky notes allover the screens, dual screens
with like Google sheets behindwith all sorts of information
because of the complexity ofscheduling a physician
(10:47):
appointment.
It's not like going and gettingyour hair done.
People are like,"Why can't weautomate more patient
scheduling?" That's a, a verylayered discussion.
There's lots of reasons,probably some that really aren't
valid and need to be changed.
And that is our biggest blockerto being able to really adopt
(11:07):
more digital technologies, etcetera.
You know, if you look at otherindustries, when was the last
time you talked to a human in abank?
And that's a highly regulatedindustry, right?
So we're a highly regulatedindustry as well.
There's a lot we can learnthere.
It's not the same.
There's a lot of things that aredifferent.
One of the main differences thatI always like to remind my
leadership team of is (11:26):
Why aren't people adopting more
rapidly in healthcare?
The central tenant forhealthcare is good clinical
care.
It's not customer service.
It's clinical care.
We want to make sure that thatpatient is receiving the best
clinical care possible, and thatis first and foremost our
guiding light.
(11:47):
If we can also deliver greatpatient experience, wonderful,
we can do that.
But first and foremost, it'sgreat clinical care.
So it has taken time to get to apoint where this technology has
advanced enough that they canput the limited resources that
they have to do the ancillarygoals.
And we've sort of democratizedAI now with a lot of these large
(12:08):
language models because they canbuild themselves.
That is transformational forthese organizations who don't
have the IT resources to throwat, you know, coding and complex
AI.
You know, scientific work thatneeds to be done to train them.
So it's a lot easier now tostart to consume this.
(12:30):
Now, we have a long way to gowith trust and you know, how do
we start thinking about biasesand privacy and all of these
different aspects that we haveto be good stewards of inside of
healthcare.
And that's the nice thing aboutpatient experiences because it's
not clinical care, right?
You're not giving clinicaladvice.
(12:51):
And so that is a little easierplace to start than something
like diagnoses.
And those are moving reallyrapidly too.
So as you think through all ofthese different areas, you can't
walk down a trade show nowwithout seeing AI or agentic AI
all over the place.
Hims was, every single booth hadsomething about AI and agents
(13:13):
and all of those things.
So, separating out what'smarketing speak and what's real
is also a difficult thing for alot of these organizations and
they need a lot of help.
Nathan C (13:22):
One of the things that I noticed when I was at
HLTH in October, people are veryexcited to use technology to get
the busy work out of the way ofcare.
There's so much in the way ofthe workforce, let's just make
it easy for everybody please.
But this idea of how do youtrust?
(13:42):
How do you build?
How do you use AI and agentictools in the right places that
foster trust, that giveappropriate information?
I always talk about inmarketing, automation
accelerates everything in yourbusiness, especially the
dysfunction.
AI is really good atacceleration and does things
(14:04):
that takes us months in blinks.
How do you balance, innovation,the development costs like you
were noting.
Plus, very important thingsabout trust, safety, patient
care and data?
There's so many things that youare touching and impacting.
Patty Hayward (14:21):
Yeah, for sure! And you know, I think it's
important that we understandtechnology can't solve
everything.
When you automate a bad process,you just go faster with the bad
process.
So it's very important when westart a project with our
customers to take a step backand say,"Don't just pour old
wine into a new bottle." Youneed to rethink what you're
doing and how you're doing it sothat you can actually hit the
(14:44):
goals that we've set out,because any organization that's
buying something wants to lookat a return on investment,
right?
I mean, you don't spend moneywithout understanding how you're
gonna get your money back.
And so being able to hit thosereturn on investment metrics and
goals is really important.
And in order to do that, youreally need to look at your
processes, your culture, howyou're going about embracing the
(15:06):
technology, because technologycan't fix your core
dysfunctions.
You've gotta identify them andlook at ways that you can change
your processes and your culturein order to optimize the
technology.
I used to sell robots back inthe day that would do
centralized cart fill, and so itwas a really dumb machine.
(15:28):
It would just pick things up andit scanned the barcode.
It knew what it was picking up,and it would pick it and bring
it over and drop it into anenvelope.
If you don't optimize yourprocesses around what it does
well, you're not gonna do wellwith it.
Like if you wanna make it dosomething that it's not
programmed to do or in a waythat it's not optimized, you're
not gonna have the return.
And it was pretty black andwhite.
(15:49):
We had organizations that weretransformational with it.
Got all of their pharmacists outon the floors, you know, and
really were able to change theirentire culture.
Others wanted it to just doexactly what it's always done
and they didn't get the returns.
So it's really, really justimportant to do that.
Now, AI is much smarter so youcan bend it a little bit more to
your will, but then you end upwith still a lot of issues with
(16:14):
things like, okay, I have a lotof physicians that want Gladys
to still answer their phone'cause Gladys has always
answered their phone and sheknows my routine for scheduling
and how I like to be scheduledand all of this and that.
And when everything is aone-off, nothing can be
automated.
We need strong physicianleadership, right?
We need strong support of theautomation to say, these are the
(16:37):
things guys, if we can all cometogether and figure out a way
where we can sort of optimizehow we schedule.
We've had patient portalscheduling for decades! And when
was the last time you were ableto schedule online with a
doctor?
Hardly ever.
There's a few that have adoptedit, but very, very rarely.
So it's not a technology issue,it's a how do we get folks to
(16:59):
really think differently aboutthe way that they're doing
things?
And there's some greatorganizations, I've been really
leaning in to organizations thatare seeing success and
understanding, okay, how are youguys doing these things?
And really leaning into thephysician leadership, because
that's really important tounderstand a, you know, what do
they need in order to practicewell, we wanna support them.
(17:20):
That's, that's a crucial thing.
Remember, central tenantclinical care, right?
So how do we support them?
But how do we also take cost outof the business?
We all are feeling the absoluteincrease in costs around
healthcare.
It's mind-boggling how muchhealthcare costs now versus 10,
20 years ago.
So we need to do some things torein it in.
And that might mean we need tomake some changes to the type of
(17:43):
appointments that you have anddo that in a way that allows you
to still have and honor yourspecialty practice.
But at the same time allowsautomation to occur so that the
organization can reduce theiroverhead.
It's the only way we're gonnastop this train wreck of
healthcare cost increases.
So how do we continue to do thatin a way that's collaborative,
(18:05):
transformational, and reallyembracing the new technology
that the organizations havedecided to purchase.
Nathan C (18:11):
Goodness! This is like such a masterclass in
change and transformation.
Really doesn't matter how muchtechnology or how powerful of a
technology you're layering in,right?
If it's not Supporting the coregoals of the user, if it doesn't
(18:33):
have buy-in at an incrediblyhigh level early on.
When you're looking at measuringthe impact of the work that you
do today is cost really theprimary metric or, it sounds
like you're working, a lotdirectly with doctors
potentially within systems andlarger organizations.
(18:55):
How do you help them?
I think as somebody who's intransformation, this is
important, right?
Like how do you help the peoplewho are undergoing
transformation measure theimpact of it?
And then how do you measure yourimpact within that process and
that growth.
Patty Hayward (19:11):
So I think it's twofold, right?
So we do have, of course,standard like time savings
measurements, things like that.
We call it full containmentinside of a contact center,
which means we've been able tofully answer a question from,
you know, an automatedperspective, we're not, having
to transfer to a human.
So there's a big cost savingsthere, obviously.
Speed to answer.
So you've reduced the queue sizebecause you're taking off the
(19:34):
easy questions, and now you'renot having to wait a half an
hour to talk to a human being,right?
So reduction in wait time.
Big patient satisfier, obviouslythat's a huge thing.
A lot of times you're in reallycompetitive spaces where you've
got five or six health systemsthat are competing for the
business of that patient.
And if they can't get through orcan't get an actual appointment,
(19:55):
they're gonna go someplace elsewhere they can.
So, you know, whoever has thefirst opening gets the business
type of thing.
So it's important to make surethat we can address those needs,
making sure that we can helpoptimize, how fast we can get to
that patient, work with them onscheduling, making it easy for
them to schedule and reschedule.
That's really important.
And then in the member side, ifyou think about the payer side,
(20:18):
especially for MedicareAdvantage, which you do a lot of
work with those organizations,they get ranked on star ratings
and member satisfaction is ahuge piece of that.
So being able to ensure thatthey're being answered right
away, they're being thoroughlyanswered in the language that
they can speak.
All of those different thingsare really, really important to
(20:39):
these payer organizations.
And then when you think aboutlife sciences and you think
about especially a lot of thescrutiny that they get under for
pricing and different thingslike that, how do we help their
patient access centers empowerthose patients that need to take
the drugs that they've broughtto market but can't afford.
And that's their role is to takethose referrals from the
(20:59):
physicians who want to help themand get them on therapy.
How do we keep them on therapy?
Because we know it can be lifechanging, but we also know some
of these specialty drugs are$10,000 a dose.
How can we help them afford thatand truly understand what
they're supposed to do'causesome of these drugs, if you
don't take them exactly asprescribed when they are
prescribed, they're ineffectiveand they're also$10,000 a dose.
(21:21):
So how do you go about makingsure that these folks are
really, really in tune with whatthey're supposed to do when
they're supposed to do it, andthat they're helped with the
cost?
So across those different worldsare sort of what we do and can
intersect and help, to reallyallow these organizations to do
that and look across at patientsatisfaction, healthcare
(21:41):
provider satisfaction as thesehealthcare providers are calling
in member satisfaction.
Those are high ranking needs forthese guys as well as of course,
all the efficiency metrics.
How can we make sure wait timesare lower?
How can we make sure that you'reanswering on the first call as
opposed to having them call backmultiple times'cause they get
frustrated without getting theanswer that they think they
need.
(22:02):
You know, we call that"firstcall resolution." So there's a
lot of standard metrics, KPIsthat we have in contact center.
But I've always challenged myorganizations to think a little
differently too, around how dowe think about number of
appointments scheduled?
Are we able to look at, hey, youknow, as you're starting to
create time and space for theseagents, are we answering more
(22:24):
than just the question theycalled in for?
An example would be,"I've runout of refills on my
prescription." and The agent cansay,"Okay, I can help you with
that.""You haven't been in foryour annual physical.
You need to get your mammogram,you need to get your A1C
checked." This list of thingsthat are also happens to be the
value-based care metrics thatthe health system is measured
(22:45):
upon, and that's how they gettheir bonus payments.
Or conversely, they get knockedfor not meeting those metrics
that they need.
And they end up gettingpenalized.
So how do we take them frompenalty to bonus?
So that helps their bottom lineand it becomes a value driver
for everyone because if they dothe things that we've identified
in those HEDIS measures, we knowthat they're gonna stay
(23:07):
healthier.
They're gonna stay outta thehospital.
They're not gonna have theacuity that you would have if
you just ignore.
We get a lot of peopleunfortunately, who don't do the
things that they're supposed todo, and they don't come in on a
regular basis either from fear,from just cost all these things.
How can we understand what'sgoing on here and how can we
help them better afford a drugthat maybe they're only, they're
(23:29):
supposed to be taking once aday?
Now they're only taking it everyother day.
Is that cost?
Is that side effects?
How do we know that?
That's a big thing.
Patient days on therapy is ameasure for these organizations.
So how do we make sure thatthey're doing that and that we
understand the reasons whythey're not, if they're not.
Are they having problems gettingto the pharmacy?
Should we just mail it to them?
What are the ways that we canhelp these patients do the
(23:52):
things that they need to do tostay healthier, which reduces
cost, which helps with bonuspayments, which helps with a
patient experience.
Talk about the ultimate patientexperience if you stay healthy.
Lack of health is the worstdriver of mental health
problems, quality of life, etcetera.
I think being able to help guidethese organizations to guide
their patients throughcomplicated therapies and
(24:15):
protocols and appointmentscheduling and all those things,
that to me is my ultimate visionof where we go when we get sort
of time back for these agents.
Nathan C (24:24):
I'm bought on a hundred percent.
You know this, this picture thatyou painted of from the first
time I call into a call centerto schedule an appointment or to
get a clarification on somethingI'm now kind of in the care of
this agentic support, which hasa little bit of visibility into
(24:47):
the history, can start to bethat through line that used to
be the very friendly woman whoanswers the phone, for all the
doctors and really like startsto hold onto things, be a
consistent data set through thatwhole experience.
And I love this idea that, notjust providing value at that
first engagement, but deliveringengaging, teaching back into.
(25:11):
Anything that gives doctors moretime to doctor, I think the
whole system would be thrilledwith! Absolutely! So I love this
idea of learning and growing,you kinda have to paint the
picture for a team and encouragethem to go find the path, even
when it's not all there today.
So, the show is called The GlowUp.
(25:32):
A"glow up" is a notabletransformation, or a rebirth.
We use it to talk about bigaudacious goals that you might
have in the next six months,When you're thinking about
transformation in healthcare, Ifeel like any six month goal is
audacious on its own! What sortof audacious goals do you have
for your work at Talkdesk inhealth and life sciences?
Patty Hayward (25:53):
So I'll take it into two sectors because I think
there's the technology goals,which are fun and really
inspiring right now about whatwe can do.
And then there's thetransformation, sort of like,
okay, what do we need to do forthem to be able to take
advantage in a meaningful way ofthis technology?
As we've talked about, you'vegotta have both.
I think from a technologyperspective, I think this
(26:13):
agentic AI and the things thatwe are building and doing.
It's just like nothing I've everseen as far as how fast these
things are evolving and theabilities that they have with
sort of multiple, each agentkind of has its own specialty,
just like you do with humans.
And then you've got thissupervisor agent that can then
say,"Okay, I see you're callingfor this.
(26:35):
I'm gonna send you to thisspecialty agent who's now gonna
ask you the questions." You'renot gonna see that it's not
gonna transfer, it's just gonnaautomatically do that so that it
knows which group, which thingsto ask.
And that way you're notconfusing things, you're not
getting that hallucination.
All those different things thatwe've been able to solve, and
allow those to be deployed bothin a digital way, right?
(26:55):
So typing in any channel youwant, right?
So not everybody wants to call.
I know my daughter would love tonever speak to another human
again.
She'd love to text everybody.
So then there's the, you know,my dad who wants to call
everybody and I'm probablysomewhere in the middle.
And I think that, being able tomeet people in their channel of
(27:15):
choice, have those conversationsin a meaningful way in the
language that they speaknatively without having to have
them translate.
That might cause its own set ofproblems.
So being able to translate intothe language that they are
native to and being able to,really help folks.
That for me, from atechnological standpoint, the
next six months we're gonna seesome just amazing output there.
(27:37):
So I'm really excited for that.
And then on the transformationside, I think I talked a lot
about sort of what I'm reallyfocused on today, which is
really working withorganizations to look at what
are the blockers, how do we needto transform the culture and the
processes that they've builtover all of these years that may
not serve well with automation,right?
(28:01):
As people are starting to buymore and more into like, look,
we've gotta transform.
These are our importantfinancial goals that we have to
achieve and it's important forus to ladder up and down from
C-Suite all the way through tothe end users and the
physicians, et cetera.
Everybody's gotta kind of berowing in the same boat and
that's not easy! Buy in and ofitself, but then to kind of see
(28:23):
how do we make these changes?
What are some of the things thatwe can experiment with and try
maybe in small pods that aremore open to change.
What are the ones that we canuse?
Let's try.
Fail fast because these kinds ofmodels allow us to fail fast.
That didn't work?
Okay, let's try this tweak.
Ooh, that worked! And you can doa lot more changes more easily
(28:43):
with this work than in the pastwhere it required so much coding
and resources, you don't needthat as much anymore.
So it's much easier for us to doiterations and to iterate fast
to get to that ultimate, okay,this is working.
This is working really well.
So that's kind of thetransformational work I'm really
excited about in the next sixmonths, that I'm working on now.
Nathan C (29:02):
Do you use agents in that like alignment and
discovery and process work thatyou do on the transformation
side as well?
Patty Hayward (29:15):
When you say agents, are you talking about
human agents or are you talkingabout the agentic AI?
Nathan C (29:19):
Does a agentic AI support those initiatives?
Patty Hayward (29:22):
So yes, the agentic AI, so there's the
agents that answer the phone,and then there's the ones behind
the scenes that sort of analyzepatterns and different things
that they're seeing and hearing.
I have an example of one of ourfinancial services group.
They did some amazing work andused our platform that listens
and finds patterns indiscussions and things like that
(29:44):
to detect fraud in real time.
They noticed a certain patternwhen someone was calling and
trying to take over someone'saccount, and they were able to
put together a real time alertto their supervisors when they
detected a pattern like this,that they would come in and cut
it off in real time, and theywere able to completely reduce
(30:06):
the amount of money that wasbeing stolen out of different
bank accounts, by the way thatthese guys were trying to reset
passwords, right?
So I'm trying to reset apassword.
They call in with things likecrying baby in the background,
right?
Like,"I've gotta get this doneright away.
I'm late for work, my baby'scrying." You make this really
stressful situation for theagents, and the agents kind of
comply and do it.
(30:27):
They started to see that patternand they were able to cut this
off in real time.
So it's pretty exciting to seesome of the evolutions of how
this can do it.
We know in healthcare we'reunder attack on a regular basis
from people trying to doransomware for that data.
So password reset's a big dealright now.
How do we detect and look atfraudulent ways of this stuff
(30:49):
happening that we can cut off inreal time.
How can we do things like makesure that when we do have an
agent that's having some troublecommunicating in real time or
alert a supervisor to be able tokind of come in and join and
support that.
Lots of great things that we cando in real time now, that I
think are gonna get better andbetter and better.
You talked about AI editing.
It can detect and really look atthings in really good ways.
(31:12):
You can tell it,"Eliminate allthe ahs and ums and awkward
silences and things like that."That's the way things are
working in detecting.
Because we record all of thesecalls for quality purposes,
right?
How do we go about learning andtraining our agents to do
better?
Learning and training ouragentic agents to be better?
What are areas folks are gettingstuck and they have to escalate?
(31:34):
Maybe we don't need that tohappen.
So how do we iterate?
Iteration and quick iteration isreally important in this
innovation.
Nathan C (31:41):
Oh my goodness! So many seeds there! What if I'm
having a frustratedconversation, right, sort of
unpacking your fraud example,and the agentic AI could help
the agent understand thisparticular patient is getting
very frustrated.
Could you send them some flowersor whatever that digital
(32:01):
equivalent is, or maybe justtake a break and take a breath.
The ability for AI to sort ofhelp those of us who are
managing so much detail tounderstand what's important in
the moment, and to see the otherinputs that might be driving the
problem is such a powerfulthing, to think about.
(32:23):
I
Patty Hayward (32:23):
love it because it automatically takes notes and
summarizes things for me.
I'm a talk thinker.
use my hands a lot.
Trying to write things down mynotes are terrible.
They're universally terrible.
I have like one word that I waslike, for some reason I thought
that was important at one point,but I don't actually know what
it means anymore.
My notes tend to be a lot likethat because I get really
involved in a conversation andI'm really trying to actively
listen as opposed to justwriting all the stuff down.
(32:46):
This allows us to not have to dothat, right?
It can guide more and it cangive you less programmatic
responses to things which Ithink
Nathan C (32:53):
Cool.
is really going
Patty Hayward (32:54):
to help all of that.
Nathan C (32:55):
So on each episode of The Glow Up, we do make an
opportunity for a communityshout out or spotlight to a
group that's doing work for goodor impact.
Is there a spotlight that you'dlike to share your episode
today?
Patty Hayward (33:08):
Thanks for that opportunity.
April's Parkinson's AwarenessMonth and my sister-in-law
suffers from this awful disease.
She's quite young.
She's in her early fifties, soit's pretty devastating.
The support and the researchthat's going on right now, talk
about transformation, there's alot of great things that are
going on in scientific research,so I always like to highlight
the Parkinson's ResearchFoundation.
(33:30):
I think they are incredible.
They're doing great things.
They're funding a lot of amazingresearch that I would love to
see continue.
Nathan C (33:36):
Heck yeah! Thank you so much and so timely.
Sort of switching gears a littlebit, in your work at Talkdesk
and the things that you'relooking to accomplish over the
next year or so, is thereanything that you're looking
for, whether it's partners,research, team members, et
cetera?
Patty Hayward (33:54):
I think I'm gonna go back to that partnering with
the physician organizations.
I would love to get a focusgroup around just that.
And bring some really keyphysician leaders together to
really think through how do wethink about automation in a
different way?
And how can we makerecommendations and have
multiple types ofrecommendations depending on the
(34:15):
organization that would allowboth their practice to thrive as
well as automation to continueto evolve so that there's more
and more things that we can do,overheads reduce all of those
different great things.
I'm very clear on clinical careis the central tenet.
That's really important.
We don't wanna change that.
We're not looking to take thataway.
(34:35):
And then how do we go forwardlooking from that?
So that's been my primary focusfor the last several months as
I've started to have a lot offolks go live with the new
technology and sort of start torun into some blockers of
adoption.
Primary care is pretty simple todo'cause it's a pretty standard
type of care.
But once you get into theologies and all of the different
(34:57):
types of specialties out there,and then you have subspecialties
and subspecialties, so muchcomplexity! Don't even get me
started on radiology.
You know, how do we go aboutscheduling all of these things
in a way that allows us to honorthe needs of making sure folks
have prior authorizations andall of those different things,
but at the same time, allows formore automation so that we can
(35:18):
do more and more things like wetalked about with care guiding.
Nathan C (35:21):
I love that the call to action is around building
partnerships and doing consumerresearch.
Could not be more on brand forThe Glow Up! I love it! Patty,
it has been such a delight tochat with you.
Thank you for spending so muchof your time today, with us on
The Glow Up.
If people wanted to learn moreor join in a research
(35:42):
partnership, what's the best wayto follow up with you or connect
online?
Patty Hayward (35:47):
Probably LinkedIn.
That is probably the best placeto go.
I publish all of my thoughtleadership and podcasts and
different things that we'redoing at Talkdesk there, so it's
a great place to see what'sgoing on and to connect with me.
Nathan C (35:59):
Patty Hayward, it has been such a fantastic treat for
me to talk about AI drivenhealthcare transformation with
you.
I really appreciate so much yourfocus on That central goal of
working in the healthcare space,which is provide top quality
(36:19):
care and to power, peopleproviding it to do the best,
they can with so manyconstraints and so many
challenges.
I'm really heartened to hear theAI story in healthcare moving
beyond just like,"Hey, we'll fixyour data." but We'll actually
like activate your data, empoweryou, and really start to build
(36:43):
this model for a unifiedcontinuum of care, not just like
40 different disparateconversations where you say 60%
of the same thing over and over.
I've never been so excited abouthealthcare scheduling and
supporting doctors withautomation.
Thank you so much for thefantastic work that you're doing
(37:05):
and sharing it with us on TheGlow Up.
Patty Hayward (37:07):
Absolutely! Thanks for having me!
Popular Podcasts
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.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
NFL Daily with Gregg Rosenthal
Gregg Rosenthal and a rotating crew of elite NFL Media co-hosts, including Patrick Claybon, Colleen Wolfe, Steve Wyche, Nick Shook and Jourdan Rodrigue of The Athletic get you caught up daily on all the NFL news and analysis you need to be smarter and funnier than your friends.