May 8, 2025 • 52 mins
“We want to bring health care to a place where claims are processed in real-time, like checking out on Amazon,” says Stephen Krupa, CEO of HealthEdge. Krupa joins Bloomberg Intelligence analyst Jonathan Palmer to unpack how HealthEdge’s core platform powers claims processing for more than 110 million people. He shares how legacy infrastructure still dominates (with 60% of health plans still relying on decades-old technology), why automation and gen AI will drive touchless transactions and how payer platforms must evolve to support member-centric care. Krupa also reflects on HealthEdge’s journey from startup to its transformation under Blackstone, and what the next chapter might look like with Bain Capital.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:18):
Welcome to another episode of Bloomberg Intelligences Vanguards of Healthcare podcast,
where we speak with the leaders at the forefront of
change in the healthcare industry. My name is Jonathan Palmer,
and I'm a healthcare analyst at Bloomberg Intelligence, the in
house research arm of Bloomberg. I'm very happy to welcome
today's guest, Steve Krupa, the CEO of health Edge, a
software platform serving over one hundred and fifteen health plans
(00:39):
that represents more than one hundred and ten million lives.
Prior to joining the company almost a decade ago, he
was a partner at the venture group that helped fund it,
and he's also the host of the Breaking Health podcast.
Thanks for joining us today, Steve.
Speaker 2 (00:51):
Happy to be here.
Speaker 1 (00:52):
Well, why don't we dive right in. What's the elevator
pitch for what health Edge does and kind of where
do you sit in the healthcare ecosystem.
Speaker 3 (00:58):
Well, flagship products called health ruth Payer and it sits
it is the engine that pays claims for our customers.
So I think the elevator pitch is around creating greater
efficiency and accuracy of the claim system and ultimately, you know,
our vision for the company.
Speaker 2 (01:17):
Which is a long term visions.
Speaker 3 (01:19):
We say, let's innovate a world where healthcare can focus
on people. And I think what we experience most of
the time when we enter into the healthcare system is
we find it's a system that's based on treatment and reimbursement,
and the person sort of gets lost in that oftentimes.
And the view is is that until you can create
(01:39):
these efficient processes around payments and treatments and claims and
accuracy around that side of it, and then enable that
clinical environment to take over and deliver great service to
the member. So our objective is to use technology to
enable our health plans to become more member centric and
(01:59):
more focused on the delivery of the appropriate care.
Speaker 2 (02:03):
In an efficient way.
Speaker 3 (02:04):
And by doing that, we need to automate the whole
back office. Not that different the way and Amazon experience
is automated.
Speaker 1 (02:11):
Interesting because I'm going to come back to that analogy
maybe a little bit later, sure, but maybe let's talk
about the journey of health Edge. You've been involved with
the company for twenty years twenty years. What was the
vision when it was first pitched to you back in
two thousand and four or five six around that timeframe.
Speaker 3 (02:26):
It was a little simpler from a standpoint of We
invested in the product health Rules Payer in its early designs,
and it was originally conceived by people that had operated
health plants. So their argument was, I think it works
very well today. Is the administration of new benefit programs,
(02:50):
the administration of new forms of contracting with providers.
Speaker 2 (02:53):
You know, value based, capitation, fee for service.
Speaker 3 (02:56):
Everything that's done there is very inefficient from a standpoint
of programming the system to operate in a fully automatic way.
Speaker 1 (03:04):
Okay, So we.
Speaker 3 (03:05):
Backed health Edge when it had something called health rules Language,
which basically inside the software, it has the intelligence to
understand the way benefits are structured and the way the
contracts are structured, and it enables a higher level of language,
really a special purpose language that feels, looks and feels
just like a contract might.
Speaker 2 (03:26):
So if you.
Speaker 3 (03:27):
Redesign your benefits, configuring the system to adjudicate those benefits,
it's fairly straightforward. You just translate the benefit document that
we all see when we design up for our health plans.
You al must read them, but some of us do,
and you translate.
Speaker 1 (03:42):
Listening to this podcast, definitely don't.
Speaker 3 (03:44):
Yes, anyone listening or knows that they should read them anyway,
but it translates that information into operational rules such that
claims that can come in can be adjudicated according to
those documents, and it can be audited back in real
English language.
Speaker 2 (03:58):
So you get out of program and you.
Speaker 3 (04:01):
Get into English language based on Now, remember twenty years ago,
that was quite an innovation.
Speaker 1 (04:05):
So how are they doing it? Maybe what was the
status quo at that time?
Speaker 2 (04:08):
The status quo would be.
Speaker 3 (04:12):
Branching configurations, okay, which would be like a hierarchical design
that says, I have this benefit, now I'm going to
create this sub benefit around cardiac or the sub benefit
around ortho.
Speaker 2 (04:25):
And so on.
Speaker 3 (04:25):
And it was very waterfall and very error prone, so
if you made a mistake, you sort of had to
rotate your way back to the very beginning and start again.
Whereas Health rules language, because it inherently understands benefits as
you make changes, it's able to auto adapt.
Speaker 1 (04:39):
I say language. And so, what were some of the
early successes with the Health rules language we saw customers?
Speaker 3 (04:47):
One of the most interesting case studies, which is probably
now several almost a decade old, was when Obamacare came out.
It basically eliminated a whole layer of the individual insurance
programs and substituted the Obama plan structure. So what happened
was a lot of our customers decommissioned those programs, and
(05:10):
then Obama came out and said, well, if you like
your old plan, old heel plan, you can have it.
Speaker 1 (05:13):
I'm sure I remember.
Speaker 2 (05:14):
This has made a lot of new noise and news.
Speaker 3 (05:18):
But they put our customers in a tough spot because
now they had to go back and recommission everything they
had decommission. And we were able to see at least
several customers. One that comes to mind in particular, that
was able to reconfigure about sixty health plans and have
them back in the market ready to adjudicate claims in
several weeks. In most of the legacy platforms that were
(05:40):
installed at that time, and today as well, if it's
a legacy platform, that operation would be several months in time.
Speaker 1 (05:46):
Really, yeah, maybe define the landscape a little bit better
for somebody who like myself, who's not that familiar with it.
How many of those legacy platforms are still in place today?
Is there's still technology from the early two thousand and
late nineties. Maybe keep going about how far back can
we actually get.
Speaker 3 (06:03):
Well, you can go back to in some cases the
seventies and eighties, like when I was a kid in college.
I'll date myself now.
Speaker 2 (06:13):
But like the.
Speaker 3 (06:15):
Language that were popular for business applications were Assembler and
then ultimately Cobaal and the science applications people were using
four tran and we were just starting to use basic
Apple Basic for example. A lot of the systems that
process claims today in fairly decent sized plans run on
Cobylt based software. Many of them run on Assembler based
(06:35):
software really where you would go in and you would
see like a green screen and it would have you know,
a flashing cursor and you type in your commands and
then run it that way. There are other systems, but
about sixty percent or so of the market is.
Speaker 1 (06:51):
Still in that old format.
Speaker 2 (06:53):
Well it's not.
Speaker 3 (06:53):
All it's going back that far, but certainly it's running
on systems that are of the vintage of the nineties.
Interesting and a lot of that is because the effort
to change out acclaim system for health plan is not insignificant.
Now technology is coming and is in place even today
that probably has reduced that cost down at least an
(07:14):
order of magnitude from where it might have been twenty
years ago, and I would imagine we'll cut that cost
in half over the next two or three years as
well as we start to apply geni AI type of
capabilities to the process of converting a customer to a
new and more up to date system. And then also
those GENI capabilities will have that up to date system
(07:35):
running in a much different way than it would have otherwise.
Speaker 1 (07:37):
I'm going to definitely come back to the AI. Yeah,
well a, it's maybe thinking about that more. The friction
around the swap out between systems. I mean, does that
mean a customer is pretty sticky for life? I mean,
how often do people actually change their administration?
Speaker 3 (07:52):
Well, historically it's been a long time, so so obviously
if something is still running that was built in the nineties,
that's thirty five years to fix it. Well, the demands
of the modern world are breaking these systems pretty substantially.
Those demands are rapid regulatory changes and also really the
(08:15):
need to start to really pursue one hundred percent automation
in these businesses.
Speaker 2 (08:21):
You see the.
Speaker 3 (08:21):
Numbers that have come out even from the public health
plans today, and you know, as they as they start
to really accommodate the more sophisticated programs from the government,
the drive towards impatient regular regulations. That is to say,
in healthcare, they would they used to put out a
regulation say you've got five years to comply, and then
everybody would extend that to seven And now they're saying
(08:44):
you've got six months to comply, and they extend it
to eighteen months. And if you don't have a tech
system that can handle that kind of change, you end
up manually processing things or process.
Speaker 1 (08:54):
Which to be more expensive. Yes, there's all the complexity
there and the friction interesting.
Speaker 3 (09:00):
So I mean, you wouldn't want to put it to
this way. I don't think people are dying to switch
as a statement, but there are sixty percent or so
of the health plans today are probably running on systems
that are very slowly beginning to a road underneath their feet,
and that pace.
Speaker 2 (09:17):
Will accelerate sharp.
Speaker 1 (09:19):
So what does the selling cycle look like.
Speaker 2 (09:22):
For you guys?
Speaker 3 (09:23):
Once somebody makes up their mind to switch systems, selling
cycles under a year. But we have seen people come
in and out of our sales pipeline over the last
I've been in the open nine years. We closed a
couple of deals last year that I knew those people
on my first day at work, you know. And we
have other products. Yeah, So we have other products that
(09:43):
are lighter weight. We have a prospective payment integrity product
that we can talk about. We have a care management
product that we talked about, member engagement products, and a
provider data manager product that are point solutions that can
operate alongside of legacy systems. And what often happens with us,
our customers become acquainted with our products by first going
(10:04):
in that direction and then use.
Speaker 1 (10:06):
That as a tip of the spear and then kind
of nudge them over.
Speaker 3 (10:09):
Yes, I like to say that there's no push sell
in selling a claim system.
Speaker 2 (10:14):
It's a pull sale.
Speaker 3 (10:15):
Somebody decides they want they need it, and then they
come to market and they do their investigation, and we're
fortunate enough to be selected quite a few of the
times that customers come to market. But the other products
that we offer are are really pathways to modern modern tech. Okay,
So one of the bigger problems in the claim side
(10:37):
of is the accuracy of the submission itself. So you
can adjudicate a submission and run it up against benefits
and running up against contracts, and come up with your
payment to the provider and your EOB and your exclamate
explanation of payment to the provider and really be able
to audit that trail all the way back to show
how that claim was handled. If the claim itself is inaccurate,
(11:00):
it doesn't represent the clinical performance that was undertaken, or
it's inappropriately submitted. That is to say, perhaps there's someone
else that should be paying that claim for a car accident.
It might be a auto insurance company, or there might
be a subrogated issue where that claim is subject to
(11:22):
a legal challenge, or it isn't really accurately representing what
the clinical work that was done. That is a fairly
large amount of the claims that gets submitted are in
error in that regard. Some of it's fraud obviously, but
discounting the actions of Tony Soprano and.
Speaker 1 (11:40):
Hisn so maybe putting a fine point on that, what
are some percentages or.
Speaker 3 (11:46):
Around it's around ten to twelve percent are inaccurately.
Speaker 1 (11:49):
Processed of all medical claims.
Speaker 2 (11:51):
Of all medical.
Speaker 3 (11:51):
Claims, they're either and about half of them are are
underpaid and the other half were overpaid. And of course
it's a very large business of payment recovery that exists
where companies data mind paid claims and determine that they
were overpaid in a pursuit to get to get them
to be more accurate. Underpayments are often discovered by the
(12:15):
providers themselves, and so they'll they'll submit an appeal to
be properly paid.
Speaker 1 (12:19):
And is that because they maybe didn't accurately describe the
clinical workflow that they actually undertook minimized it in some way,
shape or form.
Speaker 3 (12:27):
It's possible, well, and certainly in the over payments that's
the case. So you can things like over using code
modifiers for example, can lead to an overpayment. In the
case of underpayments, many times it is the case that
the pricing may not have been up to date. Okay,
if it's particularly if it's a regulatory product like Medicare
(12:49):
or American Medicaid, or the provider failed to submit a
fully accurate claim.
Speaker 1 (12:56):
Okay, that makes sense. So maybe going back to you know,
help is the core. Can we talk a little bit
about those other legs to the stool a little bit
and you know where we're in the evolution of health
Edge did those come into play? You know, when did
you start each one of those or acquire each one.
Speaker 2 (13:12):
Of those businesses?
Speaker 3 (13:14):
So we started to acquire businesses in twenty twenty, right
around the time of pandemic. We were acquired by Blackstone
at that point in time, and we principally had our
health Rules business and we had a small care management
application that was mostly you know, focused on utilization management.
And then we acquired a business called the Burgess Group
and a product called Source. So health Edge Source is
(13:37):
a prospective payment to ed grea engine that prices claims
based on regulatory pricing structures, can edit claims. That is
to say, it can pull together a set of codes
and organize them correctly. Sometimes that may mean the elimination
of some codes, It might mean the addition of some codes,
but or it may mean the adjustment of some And
(13:57):
it also enables our custom to create their own either
alerts or edits around claims based on their experience with
recoveries or underpayments. So the whole idea is to have
a pre payment mechanism that can discover some of those
errors that we talked about before they manifest themselves into
a business expense. So if you underpay a claim, you've
(14:20):
got to have a whole staff of people that handle appeals.
Speaker 2 (14:23):
And if you overpay a claim.
Speaker 3 (14:25):
You have a whole workstream that goes about trying to
discover that.
Speaker 1 (14:29):
And are those processes typically manual?
Speaker 3 (14:32):
They are machine assisted processes, so there will be a
manual element to it. I mean, the recovery business is
built around auditing essentially, so you audit claims, so you
can use machines to help with the auditing process. Ultimately,
it will touch either a clinical person or an administrative
person to determine whether any adjustment that might be pursued
(14:55):
is in fact appropriate. But generally speaking, what happens when
you edit a claim is it's edited in accordance with
medical policies that are published by the plans or by
the government, and that you can then point to and say,
I adjusted your claim this way because of this policy,
and plans promulgate policies at a pretty high rate as
(15:18):
they begin to discover the way in which they want
certain things to get reimbursed.
Speaker 2 (15:22):
Interesting give you a simple example.
Speaker 3 (15:24):
You see a cardiologist, for example, the basic cardiology exam
will be an EKG and then an ultrasound of your
heart and your kidney is and your a order. Right,
that's the first step to being prescribed to stress test,
which is the first step to being prescribed you know,
a calcification study, whether that be imaging or however they
(15:48):
want to do it. So that's an example of the
medical policy. Right, So if you're building someone for a
stress test, the first thing that you go back and
you get all that since claims history and you say, well,
did they have the appropriate exam that led to that
and if they didn't, that claim would be rejected.
Speaker 1 (16:08):
Interesting, Okay, fascinating. So that's the that's the payment integrity side.
But you've also acquired a couple other businesses as well.
Speaker 3 (16:15):
So if you think about health plans today, particularly when
they're in the government business, so that could be the
public exchanges, Medicaid or Medicare, they're not doing what they
would ordinarily do in.
Speaker 2 (16:27):
The commercial business.
Speaker 3 (16:28):
So ordinary Bloomberg would submit all their claims data to
a health plan and they would underwrite the cost to
the liver healthcare in the subsequent year based on some
documented history of what the cost had been previously. So
they get to really assess the risk before they price
the risk, and the government products health plans are assessing
(16:52):
risk based on community standards. But whoever chooses their plans
could be either super healthy or not selection by there,
so they don't know they're underwritten pool that would be
the equivalent of the Bloomberg pool until the day they
actually look at their enrollment, which it might not be
until the end of January, say, and given that calendar
(17:12):
after open enrollment. So what health plans have now moved
from is businesses that exclusively underwrite risk to businesses that
have to manage risk. And the tool that they begin
with to manage that risk is to claim, because the
claim is an abstraction of the medical record. So we
have a care management business. All of our customers have
nurses on staff and physicians on staff, and the purpose
(17:35):
of that business is to assess the overall population that's
at risk and to identify as quickly as possible, either
emerging risk or existing risk, that being patients that seem
to be on their way or in the midst of
consuming a lot of healthcare around us. If you're clinical event,
that could be excessive chronic illness, or it could be
an acute event, one of which is good like praternity,
(17:58):
would be a one underful pointical event, right. There aren't
many other ones, but that would be one and one
silver lining, right, and everybody should applaud that.
Speaker 2 (18:09):
But the goal would be.
Speaker 3 (18:10):
To find The goal is to find those patients and
then to interact with those patients and understand their needs
and catch them in that vulnerable state. When you're in
that state, you are being marched around the healthcare system.
The healthcare system is in sort of working for you. You're
working for the healthcare system, and that leads can often
(18:34):
lead lead to inefficient care and it can also lead
to non compliance by the patient. So health plans are
incentive since they're managing risk to find those patients and
help them get the best care they can. They're also
incentive to discover that risk because they can get risk
adjustment reimbursements for the premiums that they paid and change
their underwriting profile by doing that. So that's the care
(18:56):
management side of the business. And then the member engagement
side of the business that we focus on is similar
to that, but using mobile devices and care programs that
members can follow on their own, as.
Speaker 1 (19:10):
They say, to empower the patient as opposed to going
from the other angle right.
Speaker 3 (19:14):
And these are supported by the plan and they are
really clinical, robust guidelines. They're not simple around diet and
exercise or anything like that. They're around this is how
you should take your meds. You to new your medicine.
You need to go to this doctor, you need to
understand what these blood what this blood might may be
saying in terms of cholesterol or any other other indicators.
(19:37):
And it's really designed around making that an efficient experience
and kind of have the health plan sort of working
for the member. It's one of my favorite products that
we have is it is a new product. And so
the other three are more or less got to have
products for the plans. And then the plans that want
to make a difference either in their medicaid submissions or
(19:58):
their medicareusiness in terms of star ratings, etc. Are starting
to figure out ways to engage these members and give
them more of a consumer experience.
Speaker 1 (20:06):
Because that can ultimately help that star rating, which obviously
is good fun changes reimbursement.
Speaker 2 (20:11):
So there is some incentive to do that.
Speaker 3 (20:12):
But I would tell you I think that this type
of a service will be mandated in the government products,
and then once that happens, the consultancies that advise the
commercial companies around what should be in their health plans
will jump on board as well. And some of the
blues plans in our customer base are using this as
a differentiated element in their products. For commercial Yeah, it's
(20:37):
very successfully impact.
Speaker 1 (20:38):
Well, so I want to go back to the care
piece because our care management piece because as a public
markets guy, you know CBS had some hiccups last year, Yes,
the Medicare side, huh where recently United Healthcare has had
some hiccups. There is that product really helped to design
and to help surface those issues before they become a problem.
Speaker 3 (20:58):
If we're being honest, I think some of the problems
that the public companies have had were inevitable or the misreads.
Speaker 1 (21:05):
The writing members up in their pricing.
Speaker 3 (21:06):
And well, I think if you did a survey of seniors,
for example, I think you would find that many of
them have put off some of their healthcare procedures since
COVID and they're sort of just getting started again last year.
So some of those plans did not correctly calculate what
(21:27):
their utilization would be on the basis of that phenomenon.
You had a bunch of latent need that sort of
began to emerge in the marketplace in terms of being
wanting to see patients, and that was particularly acute in
the older population that was reluctant to go out of the.
Speaker 2 (21:43):
House due to exposure. We'll believe it or not.
Speaker 3 (21:47):
In twenty twenty three and twenty twenty four, they started
to get out of the house, started to feel more confident,
and as a result of their utilization and medical services
increased above what the trend would have indicated that it
would have been. I think that was a primary issue. However,
much of care management today is focused on regulatory compliance,
and there is an opportunity for those plans to drive
(22:09):
it further towards cost savings and better outcomes for members.
So they are doing it, but they're doing it inside
of a regulatory umbrella, and I'm starting to see them
think and desire to break out of that and use
it more in a proactive way.
Speaker 1 (22:24):
Is that from just an analytics perspective and thinking along
the lines of a population health.
Speaker 3 (22:29):
Yeah, okay, yeah, And you know in stars and in
medicare ratings, you have to demonstrate gaps and care closure,
So you need a care management system to do that.
You have to be able to assess the member, and
then over the course of the one year that you'll
be rated, you have to be able to demonstrate that
if you have a member, for example, that is high
(22:49):
blood pressure, that they are now on high blood pressure
medication to control that, and that it is under control.
And if you can close that gap, you get points
towards your scoring. I think more importantly, though, is the
rising risk question, which is less evident in data right away,
which is an analytics tool which says there's data that
(23:11):
suggests that Steve is about to get sick or metabolic
disease or cancer for that matter, depending on what my
lifestyle habits are, and creating those interventions earlier upfront, and
there's always a question for the health plant as to well,
how long am I carrying that risk for, as to
how much efforts should I put into that, And as
(23:32):
we start to see in Medicare in particular, there's some
brand affinity to the Medicare advantage plans. By the way,
those are very good products. They're much better than meds
up deefa service Medicare in terms of the cool the
range of services that are underwritten. There there is now
consideration towards trying to jump in and capture that rising
(23:55):
risk function before that patient becomes vulnerable inexpensive.
Speaker 1 (24:00):
Right, that's fascinating. I want to rewind a little bit
back to the Blackstone transportion. How did that change maybe
the mandate of what you were being asked to do
or how you were at being asked to run the company?
Speaker 3 (24:11):
Well, I think initially we were driving the company around
a growth curve, and we did three acquisitions very quickly.
We bought Burgess, we brought out brought out Tryst, and
we bought well.
Speaker 2 (24:23):
Framed, and those are the three digittional products.
Speaker 3 (24:26):
And then we developed internally a provider data management product,
which is a small footprint today but showing a lot
of success actually early. I think the way that what
Blackstone enabled us to do, being a startup and really
living off of very small amount of capital, was to
expand our our horizon in.
Speaker 1 (24:46):
Terms of what we capture opened up and it.
Speaker 3 (24:48):
Opened up and we really went very quickly into the
most significant areas that we felt we could have an
impact and areas where we thought our customer relationships would
enable us to expand into those areas, and that completely
changed the way the company was being rung. I was
really running the company as a former venture capitalist that
(25:10):
jumped in to run a business and was trying to
get it to an exit for the venture fund. And
it was actually a really fun experience to do that
and watch a business grow and then to have a
capital partner that was willing to work with you to
strike you know, good deals both for buyer and seller,
and then to create a commercial go to market motion
(25:33):
that enabled us to expand our impact for our customers.
Speaker 2 (25:36):
It was pretty cool.
Speaker 1 (25:37):
So now you've got the next leg with and I
know it's still early days with Baine. How does that change.
It's like a couple of weeks.
Speaker 3 (25:43):
Away, we haven't even closed it, but we signed but
not closed. Well, I think it's I think it's a
fresh legs phenomenon. I think the private equity capital cycle
is a five year cycle.
Speaker 2 (25:55):
We were right on that exactly.
Speaker 3 (25:58):
Cloth with Blackstone in April twenty twenty and you know,
we signed with with Bain in the middle of April
of twenty twenty five, so we really had five years there.
So I can be proud of the fact that stuck
around for the whole cycle with Blackstone.
Speaker 1 (26:12):
It was a lot of fun Well's even through a
couple of cycles.
Speaker 3 (26:15):
Yes, and my expectation is that we have some pretty
pretty robust plans in terms of the application of modern
technology alongside the new capabilities of this rev gen one
of AI Gen AI around really three ideas and then
(26:36):
a fourth sort of cherry on top. And first is
touchless transaction processing. So Amazon doesn't do a lot of
touching when it comes to processing your data daily needs.
Speaker 2 (26:47):
When you order.
Speaker 3 (26:47):
Books or groceries or whatever, those transactions get processed by
the machines. Ultimately the film it is human and even
the fulfillment is fairly well automated as we know, But
touchless transaction processing. Right now, auto adjudication rates Who's the
way of measuring the flow of claims through health plant
(27:09):
averages in the eighty percent range. Some of our customers
run in the low nineties. There's really no reason why
that shouldn't be one hundred percent. And the only reason
why it's not is the confidence of the rules that
the machines operate on to get the answer right. And
so you have human intervention that steps in and says
well we need to.
Speaker 2 (27:27):
Look at some of these things.
Speaker 3 (27:29):
Well, we're going to be able to do that with
jen Ai, and you're going to be able to get
these two one hundred percent. Now, once you get to
one hundred percent transaction touchless transaction processing, it changes the
whole business. It becomes a real time experience. Like how
many times do you get a bill from a physician
that's two months after You've got to service every time, right,
every time, and you're always like, well, wait a second,
(27:51):
I thought that was covered, you know, and then you
got to correlate it to your EOB and figure out
that it's been adjudicated correctly. There's no reason why that
couldn't be done within a day or so and ideally
at some point in the future at the point of service.
So you'd walk up and as you were leaving from
whatever your procedure, they would say, your copey is forty
dollars and your additional costs are going to be X
(28:12):
pay it. You'll get that weird bill. And guess why
that weird bill that you get has got a very
low collection rate. It's probably sixty five percent on the
part of the providers. Most of that stuff doesn't get paid,
so that changes the out the whole economics of the
business just in and of that. The second piece is
this we talked about, is the elimination of inaccurate claim
(28:32):
in the middle where it's one hundred percent accuracy are
ninety nine point nine percent accuracy of claims, and again
where there are gaps where we're not discovering that early on,
we can start to see Jennai having an impact on that.
And then thirdly, it's what I would call three sixty
care management, almost like care integrity, where you're doing your
(28:52):
your assessments and finding the people in the population, you're
engaging them through mobile tools, and then you're discovering your
entitlement to risk adjustment based on the average underwritten population relative.
Speaker 2 (29:06):
To what the cause for that.
Speaker 3 (29:09):
And if you pull all that stuff together, you get
sort of the cherry on top, which is end to
end business intelligence, and all of the companies that are
able to run in an end to ambitious intelligence way
run more efficiently, less expensively, that prices don't go up
some cases like AWS, prices come down. You know, it
becomes more and more efficient and as you know from
studying healthcare, medical inflation has been sort of like a
(29:32):
very painful reality for decades, we've never really gotten our
arms around that. So I think that's what the Bane
transaction would be. It will open sort of the third
book of the trilogy from my standpoint, you know, Venture Blackstone, Bane,
and that book is so it's interesting.
Speaker 1 (29:51):
You can't see it listeners. But on my little list
of questions here, one of them was, you know, will
we ever get to our healthcare experience being like our
Amazon experience?
Speaker 2 (30:00):
Yes?
Speaker 1 (30:01):
And I think the way you wait it out the
kind of points to that. So maybe the question to ask, though,
is from a health edge perspective, what sort of investments
in technology services engineers do you have to make to
actually get there? And how hard is that lift and
how competitive is it? Because I can't imagine that you're
not the only one in the space thinking this way.
Speaker 3 (30:22):
Well, you know, I think you're probably right. I think
a lot of people. I don't think there's anybody that's
listening right now that isn't nodding their head in.
Speaker 2 (30:29):
Favor of what I'm discussing.
Speaker 3 (30:31):
And then when you then try to look at where
the system and the state of the system is, what
you realize is there are a lot of very good
businesses that have been built around the inefficiencies of the system. Right,
so as you begin to erode away those efficiencies, you
impact those businesses.
Speaker 2 (30:48):
You know.
Speaker 3 (30:48):
It's a little bit like the Blockbuster effect, Like when
you apply technology, you all of a sudden, you know,
go into the video stores and what you want to
do anymore? Right, you want to go to Disney Plus
and get whatever you want whenever you want it. Right.
I think that's really the ce change that's about to
happen in healthcare is the power of GENI And this
(31:08):
is you know, the worst GENAI well play with as operators,
but that technology lay it on top of modern software,
rules based software that's auditable, that can show where everything
is taking place, so you know, gentic elements on top
of that close those edge cases. So if you said
to me today, I have this edge case where one
(31:29):
percent of my claims aren't getting adjudicated accurately because the
following variables. The expense to write that software is probably
higher than the value to the.
Speaker 2 (31:39):
Business to write the software. But if I can begin to.
Speaker 3 (31:43):
Use AI agents to create that software, create it its
auditability and sort of close that gap, I can distribute
that to customers in a very cost effective way, and
get paid the incremental value for the development support of that.
That's what changes our whole business model completely. It also
(32:04):
allows us to then demonstrate best practices in a quantitative way.
So if you configure your system in this way and
you use these agents, your accuracy delate will go up.
We can measure it, and that will drive a lot
of the adoption of new technologies. That's sixty percent that
we talked about earlier in discussion. Those people will say, Okay,
(32:27):
I've now got to do this, and that will benefit
the whole system, and we'll start it'll start to look
a lot like Amazon, and it will impact some of
the businesses today that should be doing this.
Speaker 2 (32:37):
They're going to have to do it.
Speaker 3 (32:38):
They're going to have to make a decision to keep
what they're currently doing, because it's important to be closing
those gaps of inefficiencies, but then looking at the newer
solutions to get there.
Speaker 1 (32:49):
Does anything come to mind, Maybe you don't want to
say this, but you know, where do you think the
area that's most ripe for targeting some of those inefficiencies
that are or what businesses or well subsectors or industries.
I mean, we spend a lot of time here thinking
about what AI means for you know, every subsector under
the sun, whether it's tech or consumer or you know,
(33:10):
within healthcare, pharma providers, service providers, it's it's there's everything.
Speaker 2 (33:17):
Well there there are.
Speaker 3 (33:18):
There are two components in the in the world where
I operate, which is not necessarily clinical. It's administrative with
a clinical sort of adjunct to it. But any business
that is reliant on inaccurate claims to make a living
will ultimately need to become a business that makes claims accurate, right, So.
Speaker 2 (33:40):
That'd be one example.
Speaker 3 (33:42):
Any any business that relies on aggressive billing will probably
run into trouble.
Speaker 1 (33:53):
Yeah, they'll have to take some sort of haircut.
Speaker 2 (33:55):
Because being aggressive un billing probably won't work. See.
Speaker 3 (33:58):
One of the things that happens in in the claim's
workflow is you get thirty days to process that claim
and then you run into penalties and interest that can
be very expensive depending on what state you're operating in.
So if your system is old and it's running slowly,
maybe it takes you fifteen days to do one turn.
Like you cat, you bring it in, you run it through.
(34:19):
By the time you get your answer, it's fifteen days later.
In touch less transaction processing, you have your initial answer
in seconds, right, So then you really do get a
full thirty days to make sure everything's right, and you'll
have technologies that won't need that much time, only maybe
a couple of hours more, right, So you'll so living
(34:42):
off of aggressive submissions will be hard to do, and
living off of correcting errors will become harder to do.
Is that going to happen in a snap of a finger. No,
It'll be several years for that to happen.
Speaker 1 (34:58):
That was my next question. If you think about the
different waves, whether it's the I mean, I guess we're
in the geni wave where.
Speaker 3 (35:04):
Jen Ai as a as a mind stimulus but not
necessarily an outcome stimulus.
Speaker 2 (35:10):
Ok.
Speaker 3 (35:11):
I think Jenai is an outcome stimulus is twelve eighteen
months down the road.
Speaker 1 (35:15):
And what about the agenic wave?
Speaker 2 (35:18):
The aogenetic wave is part of that. So what you know.
Speaker 3 (35:22):
I was sitting down with one of my engineers the
other day and it is a very interesting phenomenon like
and he showed me his prompts for jen Ai, and
his prompt was a couple pages long. But in that
as I sort of read through the prompt it was like,
do this, do that, calculate this, calculate that, et cetera.
I said, what you essentially have here is a programming language.
(35:43):
It's just English.
Speaker 2 (35:45):
Right.
Speaker 3 (35:45):
You're directing the agents to act like people through the
communication tool that people use, which is an English language.
So you're not using Java, you're not using objects. You
still have to access a database, right, and you still
have to host it, and you need all the processing
power to use it.
Speaker 2 (36:05):
But it is it is.
Speaker 3 (36:07):
That person that wrote that set of prompts is the
first step in AI's writing their own prompts or systems
reading their own promps, right, And once that happens, the
speed of innovation becomes incomprehensibly fast.
Speaker 2 (36:25):
Right. So we're watching these.
Speaker 3 (36:27):
Programmers sort of go from a world where maybe they
had machines that would finish their code for them or
write a unit test for them, to actually a world
where you can describe in an English language what you
want the code to do, and they can generate the
code for you. That's the beginning. In the future, that
(36:47):
code won't that necessarily always exists. You'll have an AI
tier a jen AI tier above what we think of
as an application tier, and that JENI tier will run
the application tier and eventually subsume it.
Speaker 1 (37:01):
Yeah, it'll all.
Speaker 3 (37:02):
Compress it, It'll all COMPRESSI it's just one thing, right,
more more, more, more cohesive.
Speaker 1 (37:07):
Maybe just switching gears here. We've talked a lot about,
I guess maybe the technology innovation, the business innovation. When
you think about the regulatory environment, how how is what's
happening in Washington today, And I know that's a very
fluid question. How does that change your view of where
we're going as an industry? And is there anything that
the policy makers can do to help speed this along?
Speaker 3 (37:30):
Yeah, the policy the policy makers, frankly, have historically been
the impetus for better quality health insurance, less so than
the commercial market. And I think I think that that
we will see this evolve into more of a real
pure specification writing effort in the same way that the
(37:53):
Department of Defense writes a specification about a weapon system.
UH health, Health and Human Services need to write specifications
about how they want their insurance to operate, and it
will put in the hands of the insurance companies the
onus to provide the accounting and the audit ability to
demonstrate that they've met those specifications. And so I think
(38:16):
what you'll see is, over time regulations will feel less
like a regulation and more like a specification.
Speaker 1 (38:23):
Okay, if that makes sense, No, it does. So it's
more of a mandate.
Speaker 3 (38:26):
And almost like an okay are to use, like an
internal code like companies run on goals or okay ares,
where we say we want to meet this objective and
it's this data that tells us that objective is being met.
That's more of a speculator a specification orientation, and to
say you have to do these things to demonstrate that
your compliant.
Speaker 2 (38:46):
It's totally different.
Speaker 3 (38:48):
And I think you want to spend more time focusing
on results than focusing on action. Let the plans figure
out what they need to do and then reimburse them
accordingly they have results, pay them well and if they
have average results, by them average. If they have below average,
source to pay them less. So, and I think that
(39:09):
those specifications will be extremely fluid. That is to say,
a regulator can have an idea for a new specification
and put it out and demand to go into place
within a plan year.
Speaker 1 (39:21):
Oh really that quick?
Speaker 3 (39:22):
Well, as long as the tech can handle it. Right now,
the tech can't, but the tech will be able to
handle it and not too distant future, and that will allow.
Speaker 1 (39:31):
And that'll spur all that behind.
Speaker 3 (39:32):
Bet our experimentation, broader feedback loops, et cetera. I mean,
we really will live in a world and we already do.
We just don't realize it where the government funds pretty
much all healthcare anyway. I mean, you're getting your tax
deduction right both to the employee and the corporation on
the commercial and then you're getting pretty high subsidization of
Medicare and Medicaid and a very meaningful, meaningfully large population
(39:56):
that's on Medicaid. And I know they're talking about cuts,
but I haven't really heard anything about I got about that.
Speaker 1 (40:01):
That's for a different guest.
Speaker 3 (40:02):
But over time, since how since the government is funding it,
that they're going to be the biggest influence or in
this march forward. And what we need to hope is
that the people that are put in these positions understand
the power of the new technology and what it can
do to simplify what they do but also deliver better
results than what they're getting today.
Speaker 1 (40:21):
Do you see any signs or green shoots from this
current administration that they have that they're thinking about it
in that way.
Speaker 3 (40:28):
I think the current administration is operating on with a
reform agenda from what I can tell, and whether you're
a grew with a reform or not, that seems to
be what their approach is. So in my mind, they
have to get through their reformation before they can start
to think about what they proactively want this to look like.
Speaker 1 (40:45):
Interesting maybe switching gears here. And you know, I'm an
analyst and I would be remiss if I didn't think
about some of the business KPIs. You know, what can
you share from that perspective, whether it's revenue like, what
is what does your revenue model look like?
Speaker 2 (40:58):
How do you guys?
Speaker 1 (40:59):
I didn't ask even how you get paid?
Speaker 3 (41:02):
We get paid on a SaaS fee basis. Primarily about
twenty five percent of our businesses professional services. Some of
these products, particularly the claim side, comes with a very
meaningful implementation, So we do professional services for our customers
in that regard and we work with them post implementation
to optimize that system. So that's about twenty five percent
(41:22):
of our revenue. I think they've published some anecdotal information
in the paper about our financials, which are probably not
at liberty to tell you specifically about, but I would
say that those are close to being true. Yeah, it
was like this was a pretty good reporter. They got
pretty good information from somebody. But on the software side,
we generally follow the underwriting model that our customers follow,
(41:45):
which is they built premiums on a prember per month basis,
so we charge per member per month fees for most
of what we do.
Speaker 1 (41:53):
Is there risk element at all?
Speaker 3 (41:54):
A little bit on the payment integrity side. If we
are enabling the recapture of dollars prior to payment that
would otherwise have gone out, there's there's sort of an
incentive for us to do that. We get an incentive payment,
which feels like it is almost like a risk component.
It's not quite at the level of those payments that
with the recovery vendors would be getting, but it's just
(42:17):
very very small portion of our revenue.
Speaker 2 (42:18):
Seventy five percent of our.
Speaker 3 (42:19):
Revenue is PMPM fees, less than one percent today is
based on these incentive payments, and then the other twenty
five percent professional services excellent.
Speaker 1 (42:29):
You know, maybe thinking about your journey. I mean, you
started off as a mechanical engineer banking, then went to
the VC world. You talked a little bit about, you know,
the reason you joined health Edge. I think one was
to get a transaction done. But maybe can you talk
about that journey and maybe what's what have you really
enjoyed about being the CEO of a company as an
operator as opposed to just writing the check as the VC.
Speaker 3 (42:50):
Well, you know, I think what you realize, you know,
And I was a mechanical engineer. I worked for Johnson
Controls for five years and I was really like, I
did a lot of cool stuff there.
Speaker 2 (42:57):
I did project.
Speaker 3 (42:58):
Management, I did so for engineering, I did sales, I
did a bunch of stuff. And then I went to
business school. And back then in the nineties, you got
sucked into Wall Street. So I went into M and
A for a while, and that's how I became a
healthcare person. I thought i'd be you know, energy or telecom,
because that's really what my background would have suggested. But
the first pitch I went on was to at.
Speaker 1 (43:20):
Well you avoided en Roun and yeah, I did some
of the other ones.
Speaker 3 (43:25):
Yeah, I might have missed the early in internet days
which might have been kind of fun, and I might
have ended up living in northern California. But that's that's here,
neither here and r there. Venture was about, you know,
expressing ideas in early stage companies a very binomial distribution
of success or either super successful you're not. The one
thing that I learned over the last nine years is
I do a lot of engineering work today, Like I
(43:45):
really have to understand how we build our software, which
is an engineering process.
Speaker 2 (43:50):
But what you realize it.
Speaker 3 (43:51):
Going from really the financial services world and then the
investment world and then into a business, is that running
a business as a craft it really is. It's something
that what you're really doing is is you're organizing the
activities of a lot of people to accomplish results that
benefit your customers, your employees, and your investors.
Speaker 2 (44:13):
And when you.
Speaker 3 (44:14):
Get your head around that, you really realize that you
have to lay out a long term strategy. You've got
to pull that strategy back into the current year or
the current quarter of your public company. Certainly we're judged
on our quarters as a private company, but primarily drylaged
on our years, and then you've got to set in
motion what has to be accomplished at a results level,
(44:38):
a KPI level, in terms of your commercial activities, your
engineering activities, your customer activities, how you're building a culture
for employees so you can track the best people, how
you manage all of that. And that craft is incredibly rewarding.
It's not necessarily awarding in a day to day basis,
but it's rewarding on our looks back as fact. That's
(45:00):
a I've used this quote now in the last couple
of weeks, but I think it's attributed to Dorothy Parker,
who was a writer who said, I hate writing, but
I love having written.
Speaker 1 (45:12):
That makes sense. I can see that.
Speaker 2 (45:13):
Can you see that? The writing part it's hard. But
then when you're.
Speaker 3 (45:16):
Done and you look back on what you've accomplished, and
and I encouraged our company to do this every year,
we said, wow, we really did a good job this year.
You know, we had some headwinds here, and we had
some issues here, and we got through it, and a
lot of the things that we started the year trying
to get accomplished, we've accomplished. And then of course old
glory is fleeting and then you got to have the
(45:37):
next year. But that craft is a pretty significant skill.
And it's incredibly an incredibly interesting that.
Speaker 1 (45:43):
When you look back at you know, maybe nine ten
years ago when you when you first sat in the chair,
and you think about how you thought about things or
the framework you were thinking about things, has it changed much?
And who are you using as mentors back then? And
maybe I'm sure it's evolved your leadership style, but you know,
how do you think it's changed the most in the
last ten years?
Speaker 3 (46:02):
Well, I mean, I would tell you I was supposed
to be just an interim when.
Speaker 1 (46:06):
We had hired search firm to take the job.
Speaker 3 (46:10):
A guy named Al Waxman, who subsequently has passed away,
was my business partner at a venture group and aloud
run businesses, so I had the benefit of his wisdom.
Al's greatest quote is, you don't get what you would expect,
You get what you inspect. A little bit different approach,
but it's similar in some ways. And back then I
(46:31):
had a lot of challenges sort of coming at me
and I really didn't know what the answers were. And
I was able to bring in a couple of people
that really helped me. A guy named John Lopozona came
in and he helped me be sort of a witness,
you know, like John, I think this is what's going on,
and he would say yes, And he was an old
friend and he really.
Speaker 2 (46:51):
Helped me a lot. I read.
Speaker 3 (46:56):
I read Bennett Horrowitz's book The Hard Thing About Hard Things,
and I found that book to be quite useful from
a standpoint of understanding that it's really you in the
CEO role, Like nobody cares that you're having a bad
day and that you have to get through, and then
you have to get through all of that. So that's
early the early days. And then I had a couple
(47:18):
of board members in the Blackstone chapter, Rick Jellnik and
Jay Geller, who had really good business experience that really
educated me as to the craft of running a business,
of really understanding, you know, how to plan it and
then get into it with your groups, how to organize
the business around an operating model, and how to create
(47:41):
a people model, an organizational structure and expectations and performance
driven outcomes.
Speaker 2 (47:49):
And I think that's.
Speaker 3 (47:49):
Really what's what I'm practicing today is a refinement of
that process. That process, and they were great on the board.
I mean, I would have loved it if they would
have given me a correspondence course when I owed up
nine years ago. This is these are all the things
that a CEO needs to be of doing. But we're
it's it's totally a work in progress, and you're a better,
hopefully a better CEO each day than you were the
(48:11):
day before.
Speaker 1 (48:11):
I mean, you did a couple of acquisitions a couple
of years ago, and I'm sure integrating all that and
creating kind of a whole holistic go to markets took
a lot of your time. Where are you spending all
of your time today or maybe has it shifted in
the last year or two at all.
Speaker 3 (48:25):
Yeah, I think a lot of what we're doing today
is we're focusing on how to deliver this customer experience
and create a culture that really thinks that way. So
I think what what I think has happened to healthcare
in an unfortunate way is healthcare has sort of settled
on less than great. That's our expectation that yeah, it's
like you know me, may it is better than terrible, right,
(48:49):
But the truth of the matter is is that we
need to deliver to our customer the kind of thrill
that people get when they have an Amazon experience. So
they have that media experience, and so they have whatever
that great experience that they might have with a vendor
is because then that the next step is is that
it enables our customers to deliver that experience to the
(49:09):
providers and deliver.
Speaker 2 (49:10):
That to the members.
Speaker 3 (49:10):
And if he down, yeah, and they all want to
do it. You know, there's not any desire, a lack
of desire to have a better relationship with providers and
have members be delighted with the way their healthcare is
being financed and the way they're able to access care.
But I think that the culture that we are trying
to build is we're trying to say, hey, you know,
(49:31):
let's proactively understand the problems that our customers are having
so that we can proactively solve them through technology, as
opposed to sort of settling for some level of better
slightly better than it was well said, so, Steve, One
of the ways I like to wrap these conversations up
is with a little bit of a discussion about maybe
a life lesson or something from your business career that
(49:52):
really drives you or.
Speaker 1 (49:53):
Maybe informs the mission. When you wake up in the morning,
you know, is there one thing that comes to mind
that you think about that that really informs well, you
know why you're doing what you do and why you
love it so much.
Speaker 3 (50:04):
Yeah, I mean, I would say, and Horowitz talks about
this is this whole idea of he I think he
says people plus products equals profits, and I would say
people plus products equals performance is the way I kind
of think about it. And then I go all the
way back to the beginning, and the most informative thing
that I can say about my experience is building a
(50:25):
great place to work changes everything. It changes the way
customers feel about you, It changes the way people feel
about their work.
Speaker 2 (50:35):
It gives people a home.
Speaker 3 (50:36):
To come to and they've had a bad day, and really,
you know, business can be hard and business can be painful.
And creating a company where you look out for the
employee and have them know that their needs are being
measured and you're changing the way that you run the
business on the basis of trying to give them a good,
(50:57):
great experience while they work is probably the best lesson
I think for the modern.
Speaker 2 (51:03):
World of work.
Speaker 3 (51:05):
That culture really does drive a completely different experience for everybody.
And I like to think that people come to our
company and say, Wow, this is better than the company
I just came from. This is a nice place to work.
It's a good place to work so that when we
give them the hard jobs and we give them long,
that long objective of really doing something incredibly new and different,
(51:29):
that they can approach that with energy and enthusiasm and desire.
Speaker 1 (51:33):
Yeah, it sounds like you all get to the top
of the mountain together.
Speaker 3 (51:35):
Yes, and that's fun. You know, you can always say
to people that was fun.
Speaker 2 (51:38):
It was.
Speaker 3 (51:39):
Again, it's kind of hard, but now that I look
back on it, it was very rewarding.
Speaker 2 (51:43):
And I think if we can.
Speaker 3 (51:44):
Provide that for a large majority of our employees, everything
else kind of falls into place.
Speaker 1 (51:49):
Well, that's great. Thank you so much for sharing that.
I really appreciate it. That's Steve Krupa, CEO of health Edge.
Thank you so much for joining us for our latest
episode of the Bloomberg's Vanguards of Healthcare podcast. I'm Jonathan Palmer,
and don't forget to click your follow button on your
favorite podcast app or website so you never miss a
discussion with the leaders in healthcare innovation. Until next time,
(52:09):
take care,
Welcome to another episode of Bloomberg Intelligences Vanguards of Healthcare podcast,
where we speak with the leaders at the forefront of
change in the healthcare industry. My name is Jonathan Palmer,
and I'm a healthcare analyst at Bloomberg Intelligence, the in
house research arm of Bloomberg. I'm very happy to welcome
today's guest, Steve Krupa, the CEO of health Edge, a
software platform serving over one hundred and fifteen health plans
(00:39):
that represents more than one hundred and ten million lives.
Prior to joining the company almost a decade ago, he
was a partner at the venture group that helped fund it,
and he's also the host of the Breaking Health podcast.
Thanks for joining us today, Steve.
Speaker 2 (00:51):
Happy to be here.
Speaker 1 (00:52):
Well, why don't we dive right in. What's the elevator
pitch for what health Edge does and kind of where
do you sit in the healthcare ecosystem.
Speaker 3 (00:58):
Well, flagship products called health ruth Payer and it sits
it is the engine that pays claims for our customers.
So I think the elevator pitch is around creating greater
efficiency and accuracy of the claim system and ultimately, you know,
our vision for the company.
Speaker 2 (01:17):
Which is a long term visions.
Speaker 3 (01:19):
We say, let's innovate a world where healthcare can focus
on people. And I think what we experience most of
the time when we enter into the healthcare system is
we find it's a system that's based on treatment and reimbursement,
and the person sort of gets lost in that oftentimes.
And the view is is that until you can create
(01:39):
these efficient processes around payments and treatments and claims and
accuracy around that side of it, and then enable that
clinical environment to take over and deliver great service to
the member. So our objective is to use technology to
enable our health plans to become more member centric and
(01:59):
more focused on the delivery of the appropriate care.
Speaker 2 (02:03):
In an efficient way.
Speaker 3 (02:04):
And by doing that, we need to automate the whole
back office. Not that different the way and Amazon experience
is automated.
Speaker 1 (02:11):
Interesting because I'm going to come back to that analogy
maybe a little bit later, sure, but maybe let's talk
about the journey of health Edge. You've been involved with
the company for twenty years twenty years. What was the
vision when it was first pitched to you back in
two thousand and four or five six around that timeframe.
Speaker 3 (02:26):
It was a little simpler from a standpoint of We
invested in the product health Rules Payer in its early designs,
and it was originally conceived by people that had operated
health plants. So their argument was, I think it works
very well today. Is the administration of new benefit programs,
(02:50):
the administration of new forms of contracting with providers.
Speaker 2 (02:53):
You know, value based, capitation, fee for service.
Speaker 3 (02:56):
Everything that's done there is very inefficient from a standpoint
of programming the system to operate in a fully automatic way.
Speaker 1 (03:04):
Okay, So we.
Speaker 3 (03:05):
Backed health Edge when it had something called health rules Language,
which basically inside the software, it has the intelligence to
understand the way benefits are structured and the way the
contracts are structured, and it enables a higher level of language,
really a special purpose language that feels, looks and feels
just like a contract might.
Speaker 2 (03:26):
So if you.
Speaker 3 (03:27):
Redesign your benefits, configuring the system to adjudicate those benefits,
it's fairly straightforward. You just translate the benefit document that
we all see when we design up for our health plans.
You al must read them, but some of us do,
and you translate.
Speaker 1 (03:42):
Listening to this podcast, definitely don't.
Speaker 3 (03:44):
Yes, anyone listening or knows that they should read them anyway,
but it translates that information into operational rules such that
claims that can come in can be adjudicated according to
those documents, and it can be audited back in real
English language.
Speaker 2 (03:58):
So you get out of program and you.
Speaker 3 (04:01):
Get into English language based on Now, remember twenty years ago,
that was quite an innovation.
Speaker 1 (04:05):
So how are they doing it? Maybe what was the
status quo at that time?
Speaker 2 (04:08):
The status quo would be.
Speaker 3 (04:12):
Branching configurations, okay, which would be like a hierarchical design
that says, I have this benefit, now I'm going to
create this sub benefit around cardiac or the sub benefit
around ortho.
Speaker 2 (04:25):
And so on.
Speaker 3 (04:25):
And it was very waterfall and very error prone, so
if you made a mistake, you sort of had to
rotate your way back to the very beginning and start again.
Whereas Health rules language, because it inherently understands benefits as
you make changes, it's able to auto adapt.
Speaker 1 (04:39):
I say language. And so, what were some of the
early successes with the Health rules language we saw customers?
Speaker 3 (04:47):
One of the most interesting case studies, which is probably
now several almost a decade old, was when Obamacare came out.
It basically eliminated a whole layer of the individual insurance
programs and substituted the Obama plan structure. So what happened
was a lot of our customers decommissioned those programs, and
(05:10):
then Obama came out and said, well, if you like
your old plan, old heel plan, you can have it.
Speaker 1 (05:13):
I'm sure I remember.
Speaker 2 (05:14):
This has made a lot of new noise and news.
Speaker 3 (05:18):
But they put our customers in a tough spot because
now they had to go back and recommission everything they
had decommission. And we were able to see at least
several customers. One that comes to mind in particular, that
was able to reconfigure about sixty health plans and have
them back in the market ready to adjudicate claims in
several weeks. In most of the legacy platforms that were
(05:40):
installed at that time, and today as well, if it's
a legacy platform, that operation would be several months in time.
Speaker 1 (05:46):
Really, yeah, maybe define the landscape a little bit better
for somebody who like myself, who's not that familiar with it.
How many of those legacy platforms are still in place today?
Is there's still technology from the early two thousand and
late nineties. Maybe keep going about how far back can
we actually get.
Speaker 3 (06:03):
Well, you can go back to in some cases the
seventies and eighties, like when I was a kid in college.
I'll date myself now.
Speaker 2 (06:13):
But like the.
Speaker 3 (06:15):
Language that were popular for business applications were Assembler and
then ultimately Cobaal and the science applications people were using
four tran and we were just starting to use basic
Apple Basic for example. A lot of the systems that
process claims today in fairly decent sized plans run on
Cobylt based software. Many of them run on Assembler based
(06:35):
software really where you would go in and you would
see like a green screen and it would have you know,
a flashing cursor and you type in your commands and
then run it that way. There are other systems, but
about sixty percent or so of the market is.
Speaker 1 (06:51):
Still in that old format.
Speaker 2 (06:53):
Well it's not.
Speaker 3 (06:53):
All it's going back that far, but certainly it's running
on systems that are of the vintage of the nineties.
Interesting and a lot of that is because the effort
to change out acclaim system for health plan is not insignificant.
Now technology is coming and is in place even today
that probably has reduced that cost down at least an
(07:14):
order of magnitude from where it might have been twenty
years ago, and I would imagine we'll cut that cost
in half over the next two or three years as
well as we start to apply geni AI type of
capabilities to the process of converting a customer to a
new and more up to date system. And then also
those GENI capabilities will have that up to date system
(07:35):
running in a much different way than it would have otherwise.
Speaker 1 (07:37):
I'm going to definitely come back to the AI. Yeah,
well a, it's maybe thinking about that more. The friction
around the swap out between systems. I mean, does that
mean a customer is pretty sticky for life? I mean,
how often do people actually change their administration?
Speaker 3 (07:52):
Well, historically it's been a long time, so so obviously
if something is still running that was built in the nineties,
that's thirty five years to fix it. Well, the demands
of the modern world are breaking these systems pretty substantially.
Those demands are rapid regulatory changes and also really the
(08:15):
need to start to really pursue one hundred percent automation
in these businesses.
Speaker 2 (08:21):
You see the.
Speaker 3 (08:21):
Numbers that have come out even from the public health
plans today, and you know, as they as they start
to really accommodate the more sophisticated programs from the government,
the drive towards impatient regular regulations. That is to say,
in healthcare, they would they used to put out a
regulation say you've got five years to comply, and then
everybody would extend that to seven And now they're saying
(08:44):
you've got six months to comply, and they extend it
to eighteen months. And if you don't have a tech
system that can handle that kind of change, you end
up manually processing things or process.
Speaker 1 (08:54):
Which to be more expensive. Yes, there's all the complexity
there and the friction interesting.
Speaker 3 (09:00):
So I mean, you wouldn't want to put it to
this way. I don't think people are dying to switch
as a statement, but there are sixty percent or so
of the health plans today are probably running on systems
that are very slowly beginning to a road underneath their feet,
and that pace.
Speaker 2 (09:17):
Will accelerate sharp.
Speaker 1 (09:19):
So what does the selling cycle look like.
Speaker 2 (09:22):
For you guys?
Speaker 3 (09:23):
Once somebody makes up their mind to switch systems, selling
cycles under a year. But we have seen people come
in and out of our sales pipeline over the last
I've been in the open nine years. We closed a
couple of deals last year that I knew those people
on my first day at work, you know. And we
have other products. Yeah, So we have other products that
(09:43):
are lighter weight. We have a prospective payment integrity product
that we can talk about. We have a care management
product that we talked about, member engagement products, and a
provider data manager product that are point solutions that can
operate alongside of legacy systems. And what often happens with us,
our customers become acquainted with our products by first going
(10:04):
in that direction and then use.
Speaker 1 (10:06):
That as a tip of the spear and then kind
of nudge them over.
Speaker 3 (10:09):
Yes, I like to say that there's no push sell
in selling a claim system.
Speaker 2 (10:14):
It's a pull sale.
Speaker 3 (10:15):
Somebody decides they want they need it, and then they
come to market and they do their investigation, and we're
fortunate enough to be selected quite a few of the
times that customers come to market. But the other products
that we offer are are really pathways to modern modern tech. Okay,
So one of the bigger problems in the claim side
(10:37):
of is the accuracy of the submission itself. So you
can adjudicate a submission and run it up against benefits
and running up against contracts, and come up with your
payment to the provider and your EOB and your exclamate
explanation of payment to the provider and really be able
to audit that trail all the way back to show
how that claim was handled. If the claim itself is inaccurate,
(11:00):
it doesn't represent the clinical performance that was undertaken, or
it's inappropriately submitted. That is to say, perhaps there's someone
else that should be paying that claim for a car accident.
It might be a auto insurance company, or there might
be a subrogated issue where that claim is subject to
(11:22):
a legal challenge, or it isn't really accurately representing what
the clinical work that was done. That is a fairly
large amount of the claims that gets submitted are in
error in that regard. Some of it's fraud obviously, but
discounting the actions of Tony Soprano and.
Speaker 1 (11:40):
Hisn so maybe putting a fine point on that, what
are some percentages or.
Speaker 3 (11:46):
Around it's around ten to twelve percent are inaccurately.
Speaker 1 (11:49):
Processed of all medical claims.
Speaker 2 (11:51):
Of all medical.
Speaker 3 (11:51):
Claims, they're either and about half of them are are
underpaid and the other half were overpaid. And of course
it's a very large business of payment recovery that exists
where companies data mind paid claims and determine that they
were overpaid in a pursuit to get to get them
to be more accurate. Underpayments are often discovered by the
(12:15):
providers themselves, and so they'll they'll submit an appeal to
be properly paid.
Speaker 1 (12:19):
And is that because they maybe didn't accurately describe the
clinical workflow that they actually undertook minimized it in some way,
shape or form.
Speaker 3 (12:27):
It's possible, well, and certainly in the over payments that's
the case. So you can things like over using code
modifiers for example, can lead to an overpayment. In the
case of underpayments, many times it is the case that
the pricing may not have been up to date. Okay,
if it's particularly if it's a regulatory product like Medicare
(12:49):
or American Medicaid, or the provider failed to submit a
fully accurate claim.
Speaker 1 (12:56):
Okay, that makes sense. So maybe going back to you know,
help is the core. Can we talk a little bit
about those other legs to the stool a little bit
and you know where we're in the evolution of health
Edge did those come into play? You know, when did
you start each one of those or acquire each one.
Speaker 2 (13:12):
Of those businesses?
Speaker 3 (13:14):
So we started to acquire businesses in twenty twenty, right
around the time of pandemic. We were acquired by Blackstone
at that point in time, and we principally had our
health Rules business and we had a small care management
application that was mostly you know, focused on utilization management.
And then we acquired a business called the Burgess Group
and a product called Source. So health Edge Source is
(13:37):
a prospective payment to ed grea engine that prices claims
based on regulatory pricing structures, can edit claims. That is
to say, it can pull together a set of codes
and organize them correctly. Sometimes that may mean the elimination
of some codes, It might mean the addition of some codes,
but or it may mean the adjustment of some And
(13:57):
it also enables our custom to create their own either
alerts or edits around claims based on their experience with
recoveries or underpayments. So the whole idea is to have
a pre payment mechanism that can discover some of those
errors that we talked about before they manifest themselves into
a business expense. So if you underpay a claim, you've
(14:20):
got to have a whole staff of people that handle appeals.
Speaker 2 (14:23):
And if you overpay a claim.
Speaker 3 (14:25):
You have a whole workstream that goes about trying to
discover that.
Speaker 1 (14:29):
And are those processes typically manual?
Speaker 3 (14:32):
They are machine assisted processes, so there will be a
manual element to it. I mean, the recovery business is
built around auditing essentially, so you audit claims, so you
can use machines to help with the auditing process. Ultimately,
it will touch either a clinical person or an administrative
person to determine whether any adjustment that might be pursued
(14:55):
is in fact appropriate. But generally speaking, what happens when
you edit a claim is it's edited in accordance with
medical policies that are published by the plans or by
the government, and that you can then point to and say,
I adjusted your claim this way because of this policy,
and plans promulgate policies at a pretty high rate as
(15:18):
they begin to discover the way in which they want
certain things to get reimbursed.
Speaker 2 (15:22):
Interesting give you a simple example.
Speaker 3 (15:24):
You see a cardiologist, for example, the basic cardiology exam
will be an EKG and then an ultrasound of your
heart and your kidney is and your a order. Right,
that's the first step to being prescribed to stress test,
which is the first step to being prescribed you know,
a calcification study, whether that be imaging or however they
(15:48):
want to do it. So that's an example of the
medical policy. Right, So if you're building someone for a
stress test, the first thing that you go back and
you get all that since claims history and you say, well,
did they have the appropriate exam that led to that
and if they didn't, that claim would be rejected.
Speaker 1 (16:08):
Interesting, Okay, fascinating. So that's the that's the payment integrity side.
But you've also acquired a couple other businesses as well.
Speaker 3 (16:15):
So if you think about health plans today, particularly when
they're in the government business, so that could be the
public exchanges, Medicaid or Medicare, they're not doing what they
would ordinarily do in.
Speaker 2 (16:27):
The commercial business.
Speaker 3 (16:28):
So ordinary Bloomberg would submit all their claims data to
a health plan and they would underwrite the cost to
the liver healthcare in the subsequent year based on some
documented history of what the cost had been previously. So
they get to really assess the risk before they price
the risk, and the government products health plans are assessing
(16:52):
risk based on community standards. But whoever chooses their plans
could be either super healthy or not selection by there,
so they don't know they're underwritten pool that would be
the equivalent of the Bloomberg pool until the day they
actually look at their enrollment, which it might not be
until the end of January, say, and given that calendar
(17:12):
after open enrollment. So what health plans have now moved
from is businesses that exclusively underwrite risk to businesses that
have to manage risk. And the tool that they begin
with to manage that risk is to claim, because the
claim is an abstraction of the medical record. So we
have a care management business. All of our customers have
nurses on staff and physicians on staff, and the purpose
(17:35):
of that business is to assess the overall population that's
at risk and to identify as quickly as possible, either
emerging risk or existing risk, that being patients that seem
to be on their way or in the midst of
consuming a lot of healthcare around us. If you're clinical event,
that could be excessive chronic illness, or it could be
an acute event, one of which is good like praternity,
(17:58):
would be a one underful pointical event, right. There aren't
many other ones, but that would be one and one
silver lining, right, and everybody should applaud that.
Speaker 2 (18:09):
But the goal would be.
Speaker 3 (18:10):
To find The goal is to find those patients and
then to interact with those patients and understand their needs
and catch them in that vulnerable state. When you're in
that state, you are being marched around the healthcare system.
The healthcare system is in sort of working for you. You're
working for the healthcare system, and that leads can often
(18:34):
lead lead to inefficient care and it can also lead
to non compliance by the patient. So health plans are
incentive since they're managing risk to find those patients and
help them get the best care they can. They're also
incentive to discover that risk because they can get risk
adjustment reimbursements for the premiums that they paid and change
their underwriting profile by doing that. So that's the care
(18:56):
management side of the business. And then the member engagement
side of the business that we focus on is similar
to that, but using mobile devices and care programs that
members can follow on their own, as.
Speaker 1 (19:10):
They say, to empower the patient as opposed to going
from the other angle right.
Speaker 3 (19:14):
And these are supported by the plan and they are
really clinical, robust guidelines. They're not simple around diet and
exercise or anything like that. They're around this is how
you should take your meds. You to new your medicine.
You need to go to this doctor, you need to
understand what these blood what this blood might may be
saying in terms of cholesterol or any other other indicators.
(19:37):
And it's really designed around making that an efficient experience
and kind of have the health plan sort of working
for the member. It's one of my favorite products that
we have is it is a new product. And so
the other three are more or less got to have
products for the plans. And then the plans that want
to make a difference either in their medicaid submissions or
(19:58):
their medicareusiness in terms of star ratings, etc. Are starting
to figure out ways to engage these members and give
them more of a consumer experience.
Speaker 1 (20:06):
Because that can ultimately help that star rating, which obviously
is good fun changes reimbursement.
Speaker 2 (20:11):
So there is some incentive to do that.
Speaker 3 (20:12):
But I would tell you I think that this type
of a service will be mandated in the government products,
and then once that happens, the consultancies that advise the
commercial companies around what should be in their health plans
will jump on board as well. And some of the
blues plans in our customer base are using this as
a differentiated element in their products. For commercial Yeah, it's
(20:37):
very successfully impact.
Speaker 1 (20:38):
Well, so I want to go back to the care
piece because our care management piece because as a public
markets guy, you know CBS had some hiccups last year, Yes,
the Medicare side, huh where recently United Healthcare has had
some hiccups. There is that product really helped to design
and to help surface those issues before they become a problem.
Speaker 3 (20:58):
If we're being honest, I think some of the problems
that the public companies have had were inevitable or the misreads.
Speaker 1 (21:05):
The writing members up in their pricing.
Speaker 3 (21:06):
And well, I think if you did a survey of seniors,
for example, I think you would find that many of
them have put off some of their healthcare procedures since
COVID and they're sort of just getting started again last year.
So some of those plans did not correctly calculate what
(21:27):
their utilization would be on the basis of that phenomenon.
You had a bunch of latent need that sort of
began to emerge in the marketplace in terms of being
wanting to see patients, and that was particularly acute in
the older population that was reluctant to go out of the.
Speaker 2 (21:43):
House due to exposure. We'll believe it or not.
Speaker 3 (21:47):
In twenty twenty three and twenty twenty four, they started
to get out of the house, started to feel more confident,
and as a result of their utilization and medical services
increased above what the trend would have indicated that it
would have been. I think that was a primary issue. However,
much of care management today is focused on regulatory compliance,
and there is an opportunity for those plans to drive
(22:09):
it further towards cost savings and better outcomes for members.
So they are doing it, but they're doing it inside
of a regulatory umbrella, and I'm starting to see them
think and desire to break out of that and use
it more in a proactive way.
Speaker 1 (22:24):
Is that from just an analytics perspective and thinking along
the lines of a population health.
Speaker 3 (22:29):
Yeah, okay, yeah, And you know in stars and in
medicare ratings, you have to demonstrate gaps and care closure,
So you need a care management system to do that.
You have to be able to assess the member, and
then over the course of the one year that you'll
be rated, you have to be able to demonstrate that
if you have a member, for example, that is high
(22:49):
blood pressure, that they are now on high blood pressure
medication to control that, and that it is under control.
And if you can close that gap, you get points
towards your scoring. I think more importantly, though, is the
rising risk question, which is less evident in data right away,
which is an analytics tool which says there's data that
(23:11):
suggests that Steve is about to get sick or metabolic
disease or cancer for that matter, depending on what my
lifestyle habits are, and creating those interventions earlier upfront, and
there's always a question for the health plant as to well,
how long am I carrying that risk for, as to
how much efforts should I put into that, And as
(23:32):
we start to see in Medicare in particular, there's some
brand affinity to the Medicare advantage plans. By the way,
those are very good products. They're much better than meds
up deefa service Medicare in terms of the cool the
range of services that are underwritten. There there is now
consideration towards trying to jump in and capture that rising
(23:55):
risk function before that patient becomes vulnerable inexpensive.
Speaker 1 (24:00):
Right, that's fascinating. I want to rewind a little bit
back to the Blackstone transportion. How did that change maybe
the mandate of what you were being asked to do
or how you were at being asked to run the company?
Speaker 3 (24:11):
Well, I think initially we were driving the company around
a growth curve, and we did three acquisitions very quickly.
We bought Burgess, we brought out brought out Tryst, and
we bought well.
Speaker 2 (24:23):
Framed, and those are the three digittional products.
Speaker 3 (24:26):
And then we developed internally a provider data management product,
which is a small footprint today but showing a lot
of success actually early. I think the way that what
Blackstone enabled us to do, being a startup and really
living off of very small amount of capital, was to
expand our our horizon in.
Speaker 1 (24:46):
Terms of what we capture opened up and it.
Speaker 3 (24:48):
Opened up and we really went very quickly into the
most significant areas that we felt we could have an
impact and areas where we thought our customer relationships would
enable us to expand into those areas, and that completely
changed the way the company was being rung. I was
really running the company as a former venture capitalist that
(25:10):
jumped in to run a business and was trying to
get it to an exit for the venture fund. And
it was actually a really fun experience to do that
and watch a business grow and then to have a
capital partner that was willing to work with you to
strike you know, good deals both for buyer and seller,
and then to create a commercial go to market motion
(25:33):
that enabled us to expand our impact for our customers.
Speaker 2 (25:36):
It was pretty cool.
Speaker 1 (25:37):
So now you've got the next leg with and I
know it's still early days with Baine. How does that change.
It's like a couple of weeks.
Speaker 3 (25:43):
Away, we haven't even closed it, but we signed but
not closed. Well, I think it's I think it's a
fresh legs phenomenon. I think the private equity capital cycle
is a five year cycle.
Speaker 2 (25:55):
We were right on that exactly.
Speaker 3 (25:58):
Cloth with Blackstone in April twenty twenty and you know,
we signed with with Bain in the middle of April
of twenty twenty five, so we really had five years there.
So I can be proud of the fact that stuck
around for the whole cycle with Blackstone.
Speaker 1 (26:12):
It was a lot of fun Well's even through a
couple of cycles.
Speaker 3 (26:15):
Yes, and my expectation is that we have some pretty
pretty robust plans in terms of the application of modern
technology alongside the new capabilities of this rev gen one
of AI Gen AI around really three ideas and then
(26:36):
a fourth sort of cherry on top. And first is
touchless transaction processing. So Amazon doesn't do a lot of
touching when it comes to processing your data daily needs.
Speaker 2 (26:47):
When you order.
Speaker 3 (26:47):
Books or groceries or whatever, those transactions get processed by
the machines. Ultimately the film it is human and even
the fulfillment is fairly well automated as we know, But
touchless transaction processing. Right now, auto adjudication rates Who's the
way of measuring the flow of claims through health plant
(27:09):
averages in the eighty percent range. Some of our customers
run in the low nineties. There's really no reason why
that shouldn't be one hundred percent. And the only reason
why it's not is the confidence of the rules that
the machines operate on to get the answer right. And
so you have human intervention that steps in and says
well we need to.
Speaker 2 (27:27):
Look at some of these things.
Speaker 3 (27:29):
Well, we're going to be able to do that with
jen Ai, and you're going to be able to get
these two one hundred percent. Now, once you get to
one hundred percent transaction touchless transaction processing, it changes the
whole business. It becomes a real time experience. Like how
many times do you get a bill from a physician
that's two months after You've got to service every time, right,
every time, and you're always like, well, wait a second,
(27:51):
I thought that was covered, you know, and then you
got to correlate it to your EOB and figure out
that it's been adjudicated correctly. There's no reason why that
couldn't be done within a day or so and ideally
at some point in the future at the point of service.
So you'd walk up and as you were leaving from
whatever your procedure, they would say, your copey is forty
dollars and your additional costs are going to be X
(28:12):
pay it. You'll get that weird bill. And guess why
that weird bill that you get has got a very
low collection rate. It's probably sixty five percent on the
part of the providers. Most of that stuff doesn't get paid,
so that changes the out the whole economics of the
business just in and of that. The second piece is
this we talked about, is the elimination of inaccurate claim
(28:32):
in the middle where it's one hundred percent accuracy are
ninety nine point nine percent accuracy of claims, and again
where there are gaps where we're not discovering that early on,
we can start to see Jennai having an impact on that.
And then thirdly, it's what I would call three sixty
care management, almost like care integrity, where you're doing your
(28:52):
your assessments and finding the people in the population, you're
engaging them through mobile tools, and then you're discovering your
entitlement to risk adjustment based on the average underwritten population relative.
Speaker 2 (29:06):
To what the cause for that.
Speaker 3 (29:09):
And if you pull all that stuff together, you get
sort of the cherry on top, which is end to
end business intelligence, and all of the companies that are
able to run in an end to ambitious intelligence way
run more efficiently, less expensively, that prices don't go up
some cases like AWS, prices come down. You know, it
becomes more and more efficient and as you know from
studying healthcare, medical inflation has been sort of like a
(29:32):
very painful reality for decades, we've never really gotten our
arms around that. So I think that's what the Bane
transaction would be. It will open sort of the third
book of the trilogy from my standpoint, you know, Venture Blackstone, Bane,
and that book is so it's interesting.
Speaker 1 (29:51):
You can't see it listeners. But on my little list
of questions here, one of them was, you know, will
we ever get to our healthcare experience being like our
Amazon experience?
Speaker 2 (30:00):
Yes?
Speaker 1 (30:01):
And I think the way you wait it out the
kind of points to that. So maybe the question to ask, though,
is from a health edge perspective, what sort of investments
in technology services engineers do you have to make to
actually get there? And how hard is that lift and
how competitive is it? Because I can't imagine that you're
not the only one in the space thinking this way.
Speaker 3 (30:22):
Well, you know, I think you're probably right. I think
a lot of people. I don't think there's anybody that's
listening right now that isn't nodding their head in.
Speaker 2 (30:29):
Favor of what I'm discussing.
Speaker 3 (30:31):
And then when you then try to look at where
the system and the state of the system is, what
you realize is there are a lot of very good
businesses that have been built around the inefficiencies of the system. Right,
so as you begin to erode away those efficiencies, you
impact those businesses.
Speaker 2 (30:48):
You know.
Speaker 3 (30:48):
It's a little bit like the Blockbuster effect, Like when
you apply technology, you all of a sudden, you know,
go into the video stores and what you want to
do anymore? Right, you want to go to Disney Plus
and get whatever you want whenever you want it. Right.
I think that's really the ce change that's about to
happen in healthcare is the power of GENI And this
(31:08):
is you know, the worst GENAI well play with as operators,
but that technology lay it on top of modern software,
rules based software that's auditable, that can show where everything
is taking place, so you know, gentic elements on top
of that close those edge cases. So if you said
to me today, I have this edge case where one
(31:29):
percent of my claims aren't getting adjudicated accurately because the
following variables. The expense to write that software is probably
higher than the value to the.
Speaker 2 (31:39):
Business to write the software. But if I can begin to.
Speaker 3 (31:43):
Use AI agents to create that software, create it its
auditability and sort of close that gap, I can distribute
that to customers in a very cost effective way, and
get paid the incremental value for the development support of that.
That's what changes our whole business model completely. It also
(32:04):
allows us to then demonstrate best practices in a quantitative way.
So if you configure your system in this way and
you use these agents, your accuracy delate will go up.
We can measure it, and that will drive a lot
of the adoption of new technologies. That's sixty percent that
we talked about earlier in discussion. Those people will say, Okay,
(32:27):
I've now got to do this, and that will benefit
the whole system, and we'll start it'll start to look
a lot like Amazon, and it will impact some of
the businesses today that should be doing this.
Speaker 2 (32:37):
They're going to have to do it.
Speaker 3 (32:38):
They're going to have to make a decision to keep
what they're currently doing, because it's important to be closing
those gaps of inefficiencies, but then looking at the newer
solutions to get there.
Speaker 1 (32:49):
Does anything come to mind, Maybe you don't want to
say this, but you know, where do you think the
area that's most ripe for targeting some of those inefficiencies
that are or what businesses or well subsectors or industries.
I mean, we spend a lot of time here thinking
about what AI means for you know, every subsector under
the sun, whether it's tech or consumer or you know,
(33:10):
within healthcare, pharma providers, service providers, it's it's there's everything.
Speaker 2 (33:17):
Well there there are.
Speaker 3 (33:18):
There are two components in the in the world where
I operate, which is not necessarily clinical. It's administrative with
a clinical sort of adjunct to it. But any business
that is reliant on inaccurate claims to make a living
will ultimately need to become a business that makes claims accurate, right, So.
Speaker 2 (33:40):
That'd be one example.
Speaker 3 (33:42):
Any any business that relies on aggressive billing will probably
run into trouble.
Speaker 1 (33:53):
Yeah, they'll have to take some sort of haircut.
Speaker 2 (33:55):
Because being aggressive un billing probably won't work. See.
Speaker 3 (33:58):
One of the things that happens in in the claim's
workflow is you get thirty days to process that claim
and then you run into penalties and interest that can
be very expensive depending on what state you're operating in.
So if your system is old and it's running slowly,
maybe it takes you fifteen days to do one turn.
Like you cat, you bring it in, you run it through.
(34:19):
By the time you get your answer, it's fifteen days later.
In touch less transaction processing, you have your initial answer
in seconds, right, So then you really do get a
full thirty days to make sure everything's right, and you'll
have technologies that won't need that much time, only maybe
a couple of hours more, right, So you'll so living
(34:42):
off of aggressive submissions will be hard to do, and
living off of correcting errors will become harder to do.
Is that going to happen in a snap of a finger. No,
It'll be several years for that to happen.
Speaker 1 (34:58):
That was my next question. If you think about the
different waves, whether it's the I mean, I guess we're
in the geni wave where.
Speaker 3 (35:04):
Jen Ai as a as a mind stimulus but not
necessarily an outcome stimulus.
Speaker 2 (35:10):
Ok.
Speaker 3 (35:11):
I think Jenai is an outcome stimulus is twelve eighteen
months down the road.
Speaker 1 (35:15):
And what about the agenic wave?
Speaker 2 (35:18):
The aogenetic wave is part of that. So what you know.
Speaker 3 (35:22):
I was sitting down with one of my engineers the
other day and it is a very interesting phenomenon like
and he showed me his prompts for jen Ai, and
his prompt was a couple pages long. But in that
as I sort of read through the prompt it was like,
do this, do that, calculate this, calculate that, et cetera.
I said, what you essentially have here is a programming language.
(35:43):
It's just English.
Speaker 2 (35:45):
Right.
Speaker 3 (35:45):
You're directing the agents to act like people through the
communication tool that people use, which is an English language.
So you're not using Java, you're not using objects. You
still have to access a database, right, and you still
have to host it, and you need all the processing
power to use it.
Speaker 2 (36:05):
But it is it is.
Speaker 3 (36:07):
That person that wrote that set of prompts is the
first step in AI's writing their own prompts or systems
reading their own promps, right, And once that happens, the
speed of innovation becomes incomprehensibly fast.
Speaker 2 (36:25):
Right. So we're watching these.
Speaker 3 (36:27):
Programmers sort of go from a world where maybe they
had machines that would finish their code for them or
write a unit test for them, to actually a world
where you can describe in an English language what you
want the code to do, and they can generate the
code for you. That's the beginning. In the future, that
(36:47):
code won't that necessarily always exists. You'll have an AI
tier a jen AI tier above what we think of
as an application tier, and that JENI tier will run
the application tier and eventually subsume it.
Speaker 1 (37:01):
Yeah, it'll all.
Speaker 3 (37:02):
Compress it, It'll all COMPRESSI it's just one thing, right,
more more, more, more cohesive.
Speaker 1 (37:07):
Maybe just switching gears here. We've talked a lot about,
I guess maybe the technology innovation, the business innovation. When
you think about the regulatory environment, how how is what's
happening in Washington today, And I know that's a very
fluid question. How does that change your view of where
we're going as an industry? And is there anything that
the policy makers can do to help speed this along?
Speaker 3 (37:30):
Yeah, the policy the policy makers, frankly, have historically been
the impetus for better quality health insurance, less so than
the commercial market. And I think I think that that
we will see this evolve into more of a real
pure specification writing effort in the same way that the
(37:53):
Department of Defense writes a specification about a weapon system.
UH health, Health and Human Services need to write specifications
about how they want their insurance to operate, and it
will put in the hands of the insurance companies the
onus to provide the accounting and the audit ability to
demonstrate that they've met those specifications. And so I think
(38:16):
what you'll see is, over time regulations will feel less
like a regulation and more like a specification.
Speaker 1 (38:23):
Okay, if that makes sense, No, it does. So it's
more of a mandate.
Speaker 3 (38:26):
And almost like an okay are to use, like an
internal code like companies run on goals or okay ares,
where we say we want to meet this objective and
it's this data that tells us that objective is being met.
That's more of a speculator a specification orientation, and to
say you have to do these things to demonstrate that
your compliant.
Speaker 2 (38:46):
It's totally different.
Speaker 3 (38:48):
And I think you want to spend more time focusing
on results than focusing on action. Let the plans figure
out what they need to do and then reimburse them
accordingly they have results, pay them well and if they
have average results, by them average. If they have below average,
source to pay them less. So, and I think that
(39:09):
those specifications will be extremely fluid. That is to say,
a regulator can have an idea for a new specification
and put it out and demand to go into place
within a plan year.
Speaker 1 (39:21):
Oh really that quick?
Speaker 3 (39:22):
Well, as long as the tech can handle it. Right now,
the tech can't, but the tech will be able to
handle it and not too distant future, and that will allow.
Speaker 1 (39:31):
And that'll spur all that behind.
Speaker 3 (39:32):
Bet our experimentation, broader feedback loops, et cetera. I mean,
we really will live in a world and we already do.
We just don't realize it where the government funds pretty
much all healthcare anyway. I mean, you're getting your tax
deduction right both to the employee and the corporation on
the commercial and then you're getting pretty high subsidization of
Medicare and Medicaid and a very meaningful, meaningfully large population
(39:56):
that's on Medicaid. And I know they're talking about cuts,
but I haven't really heard anything about I got about that.
Speaker 1 (40:01):
That's for a different guest.
Speaker 3 (40:02):
But over time, since how since the government is funding it,
that they're going to be the biggest influence or in
this march forward. And what we need to hope is
that the people that are put in these positions understand
the power of the new technology and what it can
do to simplify what they do but also deliver better
results than what they're getting today.
Speaker 1 (40:21):
Do you see any signs or green shoots from this
current administration that they have that they're thinking about it
in that way.
Speaker 3 (40:28):
I think the current administration is operating on with a
reform agenda from what I can tell, and whether you're
a grew with a reform or not, that seems to
be what their approach is. So in my mind, they
have to get through their reformation before they can start
to think about what they proactively want this to look like.
Speaker 1 (40:45):
Interesting maybe switching gears here. And you know, I'm an
analyst and I would be remiss if I didn't think
about some of the business KPIs. You know, what can
you share from that perspective, whether it's revenue like, what
is what does your revenue model look like?
Speaker 2 (40:58):
How do you guys?
Speaker 1 (40:59):
I didn't ask even how you get paid?
Speaker 3 (41:02):
We get paid on a SaaS fee basis. Primarily about
twenty five percent of our businesses professional services. Some of
these products, particularly the claim side, comes with a very
meaningful implementation, So we do professional services for our customers
in that regard and we work with them post implementation
to optimize that system. So that's about twenty five percent
(41:22):
of our revenue. I think they've published some anecdotal information
in the paper about our financials, which are probably not
at liberty to tell you specifically about, but I would
say that those are close to being true. Yeah, it
was like this was a pretty good reporter. They got
pretty good information from somebody. But on the software side,
we generally follow the underwriting model that our customers follow,
(41:45):
which is they built premiums on a prember per month basis,
so we charge per member per month fees for most
of what we do.
Speaker 1 (41:53):
Is there risk element at all?
Speaker 3 (41:54):
A little bit on the payment integrity side. If we
are enabling the recapture of dollars prior to payment that
would otherwise have gone out, there's there's sort of an
incentive for us to do that. We get an incentive payment,
which feels like it is almost like a risk component.
It's not quite at the level of those payments that
with the recovery vendors would be getting, but it's just
(42:17):
very very small portion of our revenue.
Speaker 2 (42:18):
Seventy five percent of our.
Speaker 3 (42:19):
Revenue is PMPM fees, less than one percent today is
based on these incentive payments, and then the other twenty
five percent professional services excellent.
Speaker 1 (42:29):
You know, maybe thinking about your journey. I mean, you
started off as a mechanical engineer banking, then went to
the VC world. You talked a little bit about, you know,
the reason you joined health Edge. I think one was
to get a transaction done. But maybe can you talk
about that journey and maybe what's what have you really
enjoyed about being the CEO of a company as an
operator as opposed to just writing the check as the VC.
Speaker 3 (42:50):
Well, you know, I think what you realize, you know,
And I was a mechanical engineer. I worked for Johnson
Controls for five years and I was really like, I
did a lot of cool stuff there.
Speaker 2 (42:57):
I did project.
Speaker 3 (42:58):
Management, I did so for engineering, I did sales, I
did a bunch of stuff. And then I went to
business school. And back then in the nineties, you got
sucked into Wall Street. So I went into M and
A for a while, and that's how I became a
healthcare person. I thought i'd be you know, energy or telecom,
because that's really what my background would have suggested. But
the first pitch I went on was to at.
Speaker 1 (43:20):
Well you avoided en Roun and yeah, I did some
of the other ones.
Speaker 3 (43:25):
Yeah, I might have missed the early in internet days
which might have been kind of fun, and I might
have ended up living in northern California. But that's that's here,
neither here and r there. Venture was about, you know,
expressing ideas in early stage companies a very binomial distribution
of success or either super successful you're not. The one
thing that I learned over the last nine years is
I do a lot of engineering work today, Like I
(43:45):
really have to understand how we build our software, which
is an engineering process.
Speaker 2 (43:50):
But what you realize it.
Speaker 3 (43:51):
Going from really the financial services world and then the
investment world and then into a business, is that running
a business as a craft it really is. It's something
that what you're really doing is is you're organizing the
activities of a lot of people to accomplish results that
benefit your customers, your employees, and your investors.
Speaker 2 (44:13):
And when you.
Speaker 3 (44:14):
Get your head around that, you really realize that you
have to lay out a long term strategy. You've got
to pull that strategy back into the current year or
the current quarter of your public company. Certainly we're judged
on our quarters as a private company, but primarily drylaged
on our years, and then you've got to set in
motion what has to be accomplished at a results level,
(44:38):
a KPI level, in terms of your commercial activities, your
engineering activities, your customer activities, how you're building a culture
for employees so you can track the best people, how
you manage all of that. And that craft is incredibly rewarding.
It's not necessarily awarding in a day to day basis,
but it's rewarding on our looks back as fact. That's
(45:00):
a I've used this quote now in the last couple
of weeks, but I think it's attributed to Dorothy Parker,
who was a writer who said, I hate writing, but
I love having written.
Speaker 1 (45:12):
That makes sense. I can see that.
Speaker 2 (45:13):
Can you see that? The writing part it's hard. But
then when you're.
Speaker 3 (45:16):
Done and you look back on what you've accomplished, and
and I encouraged our company to do this every year,
we said, wow, we really did a good job this year.
You know, we had some headwinds here, and we had
some issues here, and we got through it, and a
lot of the things that we started the year trying
to get accomplished, we've accomplished. And then of course old
glory is fleeting and then you got to have the
(45:37):
next year. But that craft is a pretty significant skill.
And it's incredibly an incredibly interesting that.
Speaker 1 (45:43):
When you look back at you know, maybe nine ten
years ago when you when you first sat in the chair,
and you think about how you thought about things or
the framework you were thinking about things, has it changed much?
And who are you using as mentors back then? And
maybe I'm sure it's evolved your leadership style, but you know,
how do you think it's changed the most in the
last ten years?
Speaker 3 (46:02):
Well, I mean, I would tell you I was supposed
to be just an interim when.
Speaker 1 (46:06):
We had hired search firm to take the job.
Speaker 3 (46:10):
A guy named Al Waxman, who subsequently has passed away,
was my business partner at a venture group and aloud
run businesses, so I had the benefit of his wisdom.
Al's greatest quote is, you don't get what you would expect,
You get what you inspect. A little bit different approach,
but it's similar in some ways. And back then I
(46:31):
had a lot of challenges sort of coming at me
and I really didn't know what the answers were. And
I was able to bring in a couple of people
that really helped me. A guy named John Lopozona came
in and he helped me be sort of a witness,
you know, like John, I think this is what's going on,
and he would say yes, And he was an old
friend and he really.
Speaker 2 (46:51):
Helped me a lot. I read.
Speaker 3 (46:56):
I read Bennett Horrowitz's book The Hard Thing About Hard Things,
and I found that book to be quite useful from
a standpoint of understanding that it's really you in the
CEO role, Like nobody cares that you're having a bad
day and that you have to get through, and then
you have to get through all of that. So that's
early the early days. And then I had a couple
(47:18):
of board members in the Blackstone chapter, Rick Jellnik and
Jay Geller, who had really good business experience that really
educated me as to the craft of running a business,
of really understanding, you know, how to plan it and
then get into it with your groups, how to organize
the business around an operating model, and how to create
(47:41):
a people model, an organizational structure and expectations and performance
driven outcomes.
Speaker 2 (47:49):
And I think that's.
Speaker 3 (47:49):
Really what's what I'm practicing today is a refinement of
that process. That process, and they were great on the board.
I mean, I would have loved it if they would
have given me a correspondence course when I owed up
nine years ago. This is these are all the things
that a CEO needs to be of doing. But we're
it's it's totally a work in progress, and you're a better,
hopefully a better CEO each day than you were the
(48:11):
day before.
Speaker 1 (48:11):
I mean, you did a couple of acquisitions a couple
of years ago, and I'm sure integrating all that and
creating kind of a whole holistic go to markets took
a lot of your time. Where are you spending all
of your time today or maybe has it shifted in
the last year or two at all.
Speaker 3 (48:25):
Yeah, I think a lot of what we're doing today
is we're focusing on how to deliver this customer experience
and create a culture that really thinks that way. So
I think what what I think has happened to healthcare
in an unfortunate way is healthcare has sort of settled
on less than great. That's our expectation that yeah, it's
like you know me, may it is better than terrible, right,
(48:49):
But the truth of the matter is is that we
need to deliver to our customer the kind of thrill
that people get when they have an Amazon experience. So
they have that media experience, and so they have whatever
that great experience that they might have with a vendor
is because then that the next step is is that
it enables our customers to deliver that experience to the
(49:09):
providers and deliver.
Speaker 2 (49:10):
That to the members.
Speaker 3 (49:10):
And if he down, yeah, and they all want to
do it. You know, there's not any desire, a lack
of desire to have a better relationship with providers and
have members be delighted with the way their healthcare is
being financed and the way they're able to access care.
But I think that the culture that we are trying
to build is we're trying to say, hey, you know,
(49:31):
let's proactively understand the problems that our customers are having
so that we can proactively solve them through technology, as
opposed to sort of settling for some level of better
slightly better than it was well said, so, Steve, One
of the ways I like to wrap these conversations up
is with a little bit of a discussion about maybe
a life lesson or something from your business career that
(49:52):
really drives you or.
Speaker 1 (49:53):
Maybe informs the mission. When you wake up in the morning,
you know, is there one thing that comes to mind
that you think about that that really informs well, you
know why you're doing what you do and why you
love it so much.
Speaker 3 (50:04):
Yeah, I mean, I would say, and Horowitz talks about
this is this whole idea of he I think he
says people plus products equals profits, and I would say
people plus products equals performance is the way I kind
of think about it. And then I go all the
way back to the beginning, and the most informative thing
that I can say about my experience is building a
(50:25):
great place to work changes everything. It changes the way
customers feel about you, It changes the way people feel
about their work.
Speaker 2 (50:35):
It gives people a home.
Speaker 3 (50:36):
To come to and they've had a bad day, and really,
you know, business can be hard and business can be painful.
And creating a company where you look out for the
employee and have them know that their needs are being
measured and you're changing the way that you run the
business on the basis of trying to give them a good,
(50:57):
great experience while they work is probably the best lesson
I think for the modern.
Speaker 2 (51:03):
World of work.
Speaker 3 (51:05):
That culture really does drive a completely different experience for everybody.
And I like to think that people come to our
company and say, Wow, this is better than the company
I just came from. This is a nice place to work.
It's a good place to work so that when we
give them the hard jobs and we give them long,
that long objective of really doing something incredibly new and different,
(51:29):
that they can approach that with energy and enthusiasm and desire.
Speaker 1 (51:33):
Yeah, it sounds like you all get to the top
of the mountain together.
Speaker 3 (51:35):
Yes, and that's fun. You know, you can always say
to people that was fun.
Speaker 2 (51:38):
It was.
Speaker 3 (51:39):
Again, it's kind of hard, but now that I look
back on it, it was very rewarding.
Speaker 2 (51:43):
And I think if we can.
Speaker 3 (51:44):
Provide that for a large majority of our employees, everything
else kind of falls into place.
Speaker 1 (51:49):
Well, that's great. Thank you so much for sharing that.
I really appreciate it. That's Steve Krupa, CEO of health Edge.
Thank you so much for joining us for our latest
episode of the Bloomberg's Vanguards of Healthcare podcast. I'm Jonathan Palmer,
and don't forget to click your follow button on your
favorite podcast app or website so you never miss a
discussion with the leaders in healthcare innovation. Until next time,
(52:09):
take care,
Host
Jonathan Palmer
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
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.
Ridiculous History
History is beautiful, brutal and, often, ridiculous. Join Ben Bowlin and Noel Brown as they dive into some of the weirdest stories from across the span of human civilization in Ridiculous History, a podcast by iHeartRadio.