Hello everyone.

Welcome back to another episode

of the Mostly Unstructured

podcast.

I'm Clay Tuten, CMO at KeyMark.

I'm joined today by a very
special guest, Mike Askren, VP

of Product at Hyland.

Mike, thanks for joining me,
dude.

Glad to see you.

It's good to be together again.

Absolutely.

You know.

A lot to talk.

About today.

Got a lot to talk about.

Yeah.

So we're gonna kick it off.

Just kind of a theme here around
the transition and the

evolution, if you will, of
document management ECM into

this intelligence layer.

Um, and so I know you've got a
lot of knowledge here.

Really glad that you're joining

me today because looking forward

to what you have to share with

us.

So are you ready to get started?

I am, let's dive into it.

Let's do it, man.

I'm going to focus today on this
topic of, um, document

management and ECM.

Right?

How that's been the table stakes
for so long.

Most organizations already have
it, but then it's moving toward

this intelligent layer and
unpacking some of that.

Right.

So, um, I think for years

enterprises have, have treated

content really.

I got, I gotta store this stuff,
I gotta secure it.

And then, you know, I, I can
retrieve it at some point.

And that's been the simple way
to put it.

Um, but we've got this
conversation change where, all

right, we know where we're,
where we're putting it, but

we're moving now to what are we,
what are we doing with it?

How are we getting intelligence
out of it?

And the real value isn't in the
file itself, but in, in what's

in the file and intelligence you
can gain from that.

So I think most organizations

are sitting on tons of

unstructured content.

You guys talk a lot about that
when you're defining

unstructured content in the
Hyland terminology, probably the

same as everybody else that, you
know, what does that look like

when you guys are talking about
unstructured content from, from

a Hyland perspective.

Yeah, it's very much the as you

said, the information embedded

within the page which which then

becomes knowledge embedded in

the page.

Um, and you're getting down to

the paragraph and sentence level

semantics of that information

and the relationships, not only

within that particular document,

but across the body of content

that it might have relationships

within the system or even other

systems.

Okay.

So, um, being able to tap into
that, um, you're certainly

hearing more terminology around
curation and vectors and chunks.

Yeah, yeah.

You know, they're, they're kind

of new terms for a lot of folks

in the industry and, um, being

able to really extract the

insight and organize it in a way

that's meaningful for this new

intelligence layer.

Yeah.

To provide value to users or to
systems driving optimizations,

automation and efficiencies.

So your, your role here at
Hyland, give me a little bit of

color around that just so that
everybody knows who's who's

giving this intelligence.

Sure, sure.

So yes, I have ownership of the

the ECM portfolio at Hyland,

both the um, core ECMs as we

might reference them of Onbase,

Alfresco, Nuxeo, Perceptive

content.

Okay.

Um, as well as our cloud
platform capability, that's sort

of delivering some of this new
intelligence layer on top of it.

Okay.

Um.

And how is that working together
with this, with the new CIC and

like just a little bit of that
would be helpful.

Great question.

So the key sort of intermediary

in that equation is content

federation.

Right.

And that is how we're providing
connectivity from these new AI

and intelligence services to the
content where it resides today,

whether that's on prem,
self-managed, self-managed in a

hyperscaler in a cloud, um, or
wherever it might exist today.

And I love the fact that Hyland
has continued to really.

And this year in the keynotes
double down on the.

The value of content.

You guys talk about content
delivery agentic.

Uh, what's your term for that

again, the content you guys are

talking about it, the agentic,

uh, enterprise.

Yeah, the content fueled agentic
enterprise.

Yeah. Sorry. I, I had it down
pat, but I thought that was a

really cool way that you guys
talked about that.

Yeah.

And then saying, hey, like AI is

great, but where the gold is

like, it's just about extracting

that.

And I think, yeah, there's this

been the evolution of ECM where

the digitization, I just gotta

digitize it.

And then kind of a second phase
is your workflows, your

automations, etc.. But now this
intelligence piece, when we say

intelligence and we throw that
term, like we all throw that

term around a lot.

So it's about as broad as cloud.

Exactly. Yeah, exactly.

So from a Hyland perspective, or
from at least this this ECM.

I don't want to say legacy isn't

a bad thing, but that's been a

heritage.

This is probably a better word.

It's ECM heritage, the value of
content.

When we're talking about
intelligence, I think it's

really helpful to explain to
people what that means, right?

That extracting meaning and, and
what else.

Like it's more than just
extracting meaning.

Like there's so much value
beyond just that.

Yes, there, there's certainly
the aspect of insights across

that information that might turn
into sort of analytics across

your content, which sort of gets
into a data analytics.

And what we've really been

promoting is for our, our ECM

owners and organizations to

start having conversations with

their data architects and their

AI architects around how you can

make that information available

to fuel, you know, not only

knowledge and awareness across

the organization, um, but also

automation concepts, you know,

built upon the years of history

of both documents and process

decisions.

Um, through that.

Yeah. Because I think, um, what
Jitesh was talking about in his

keynote, they're only able to
see fifteen to thirty percent.

It was just this stat.

Yeah.

You know, of, of the content and
just being able to expose that.

Yeah, opens up massive levels of
intelligence to that very data.

Um, and then that the agents
like, you know, we talk about

agents a lot.

Um, and the human element, we
did another podcast earlier with

the, the human element that, and
I like how you guys constantly

reiterate the value of the human
in the loop.

Yes. So I think maybe to your

first point in terms of the

amount of information and data

available in an organization,

you know, anywhere from seventy

to ninety percent of that is

still in an unstructured form

and an ECM repository that,

frankly, has not really been

paid attention to by a lot of

the other roles and functions in

an organization.

You know, roughly fifteen to
twenty, maybe thirty percent of

information in an organization
is in a structured format.

And so it's kind of untapping
this, you know, this massive

information and being able to
marry that together to generate

this enterprise context.

And, and, you know, we reference

that as our enterprise context

engine, right?

But it's not so much about the
engine.

It's about the information and

the relationships between it,

um, that can help make agents

more intelligent.

Yeah.

Such that they can reason and
understand new documents coming

into the solution and data
coming into the solution and

understands how it might relate
to an entity in a system, you

know, whether that's a citizen
applicant, a student, you know,

someone applying for services
through an organization,

whatever it might be.

Um, the tie together of that and
the context around that.

You guys had a really good job

this week of showing really

specific use cases that I think

help people kind of aha moment

like, okay, so if I'm in

healthcare, this agent can look

at, you know, my, my situation

and escalate it to get a

referral and move that through a

lot quicker.

Pulling context in from other
visits and other information.

Right.

And, and, and making that
available to the care provider.

Right.

And then just the, the speed of
that.

And then, and it, and Ed and I
were talking about this earlier.

It's so nice that it's not just
giving you the feedback, it's

also giving you context.

So this is why I pulled this.

This is where it came from.

So that you're not just having
to blindly trust it.

And I think this is where the
agentic

conversation comes in around

agentic governance, right?

Like,

so you have to we've talked
about governance on here, uh,

quite a bit on our podcast in a
broad sense, specifically
around,

you know, security, the concept
of who's got access to it,

all those good things. But

then you guys have started
really

kind of carving off and talking

about the governance of agents.

I

think that's really important to
unpack that.

Yeah, it is very important.

And it ties very it ties into

the kind of human in the loop

feedback component a little bit

as well.

Okay.

So in terms of gaining
efficiencies historically, kind

of across the solution space in
ECM, it's been about getting

the, the right document set in
front of the right person so

that they can review dozens of
pages of information, and that

might take them five, ten,
fifteen an hour to review

something really complex to get
to the decision criteria.

So if the agents in the

background can prepare that

information and decision

criteria, put it in front of the

user the moment they open the

item, the case to work, they're

able to make that decision more

effectively.

But they can also give feedback

that, hey, I believe this may be

incorrect.

Right.

And sort of thumbs down that
response, right?

You know, and that that's input
to the agent.

Yeah.

But then maybe going a little

bit further into this topic of

governance.

Um, historically in the space,
governance has had a very

specific meaning to content and
data governance, access control

combined with lifecycle stage
management of content.

And when you might ultimately
decide to disposition that or

put a legal hold on it.

Right.

Kind of the expanded realm of

that with governance is now

governance of these agents

acting on behalf of users, um,

or maybe independent of a user

context, but in the same, uh,

level of intelligence as a user,

evaluating content, evaluating

decisions and providing

recommendations.

And the agent governance layer

really provides the opportunity

to be able to, um, give control

to administrators.

We, we call that our control
tower in our particular stack.

But, you know, in, in a more
general concept, it's the

ability to monitor agents and
what they're doing and just, you

know, stop that activity.

If there's any indication based

upon various metrics that either

it's taking longer to do given

processes, which might mean

there's a new input into the

equation, you know, whatever it

might be, you think of it in the

way that you might have

monitored other systems

historically with analytics and

dashboarding tools.

The same thing for an agent and
really having oversight of them

and the ability to terminate an
agent that might be

hallucinating or might be
following the instructions that

it was given, but those
instructions may no longer be

valid or may have been provided
insufficiently, you know, in the

initial setup of the agent.

I hate to break it down too
much, but it sounds like a

digital employee in a way.

Like you have to like, if, if an
employee is not doing their job,

you have to go say, hey, that's
not actually how you do it.

They improve, they learn.

Right?

And then if that's just not
working out, then you have to

make a change.

Yes.

So I mean, I'm trying to bring a
little bit of it's coaching.

It's coaching of agents.

Yeah.

The termination is a strong
term.

Like you're terminated.

Um, but but no, I think that's I
think it's kind of helpful to

see it in that light.

Um, we interact a lot with, I'm

sure you guys do too with, um,

the CIOs, CDOs, the C-suite and

of, of, um, technology decision

makers, right.

And they have a really tough
job.

Like I was reading an article
recently where they're just

under scrutiny, like constantly.

And there's a lot of stress,
there's a lot of stress if

you're in that position.

Yeah.

Um, and I think it's important

we maybe just take a minute to

speak to them and go, hey,

listen, like here's one, two,

three things like these you

should be, you should be focused

on, right?

Like, let's walk through that to
kind of help them.

All right.

If I, because I think they feel

like they have so much coming at

them.

And that's one thing we try to
pride ourselves on is like

breaking it down.

Like, like your list.

You have a lot of voices, you're
reading articles, you're,

you're, you know, your boards
probably making demands of you.

You're just trying to pull out
some nuggets of information.

Maybe you've even deployed

something and it's not working

well.

Mhm.

Um, what have you got to say to
those folks?

I would, I would say that
everyone is facing this, this

new tool available to them and,
and technology and AI.

We've been through these hype
cycles before.

And oftentimes you have a new
technology that you're looking

for a problem to solve with.

Um, take a moment to step back
and, you know, the problems you

have in your organization.

Yeah.

And it's probably areas that

you've already been trying to

improve historically and have

been successful with that to

some degree.

Yeah.

But how can you drive more

efficiencies for humans that are

performing those activities or

further perform full automation,

you know, within a system,

within, within constraints,

within boundaries of, um, you

know, uh, limits and, and so

forth to, to allow that to, um,

occur.

So I think, you know, kind of

the trend that you're mentioning

and the pressure, um, I think

I've seen recently that, uh, you

know, the average tenure of a

CIO is something like twenty

four months, which was quite

alarming to, to me to even think

about that.

Yeah.

You know?

Yeah.

But it is sort of this pressure
for complete transformation.

Yeah.

Um, and, and I, I think there's
a real opportunity to, to make

impactful and practical changes.

Yeah.

And in adopting this new

technology and understanding the

benefits.

Yeah.

Um, what we're often seeing is

when you're trying to do a

really large, you know, complete

re-envisioning of everything

that, um, it stalls the, the

progress of that and effectively

delays the return on the

investment.

Burn.

It's just a slow burn.

Yeah.

Um, yeah, that's, that's really
good insight I think.

I, we, we like to, and I want to

see if you disagree or agree

with this.

Like we've been kind of taking

this approach, like you can do a

lot of different things, but

it's important to find the right

problem.

But don't try to.

Start with the largest problem,
correct?

Like usually if you can just get
a win.

And there's, there's actually
some pretty.

I won't say easy, but there's
some pretty.

ripe opportunities to.

And you're probably in your

processes that you don't even

know.

Right.

So sitting down with someone who

understands and letting them

just under.

You know, ask questions, find

out what you're really

struggling with.

And oftentimes, that's the
impetus that just gets the

snowball going instead of trying
to roll the big snowball in one.

You know, like you ever build a
snowman, right?

Like anybody that's done that

before knows I could kind of get

it going.

You don't start with the big
pile.

Anyways, I digress, but I think

that's important to help people

understand that.

Do you do you feel like that's a

fair assessment or would you say

no?

I think there's times where you
need to go all in.

Um, I would completely agree.

It's, it's human nature to

identify the biggest, most

complex thing that you might

actually.

What do you think that is?

You might actually do it on an
infrequent basis.

Yeah.

And perhaps that's why it's

still this big complex thing,

because it doesn't occur

frequently enough on a day to

day basis.

That's fair.

And so you haven't already tried

to improve it to a degree that

it's manageable.

So the.

Nastiest one.

You like the most, that's that's
kind of the natural thought.

It is.

But but rather than going
looking at the things that you

do most often every day that are
kind of mundane, Right.

But are there opportunities
within those to create?

You know, if you do a fifteen

minute task two hundred times a

day.

Yeah.

If you can turn that into a two
minute task two hundred times a

day, that's a pretty big time
savings every day for that

individual performing that role
and that function.

And they're more instead of
being in the role and function

of reviewing a bunch of
information, you're effectively

reviewing the key decision
criteria and making a decision.

Um, kind of back to my earlier

example, and I think that's the

real opportunity to have some

easier, more manageable to

deliver wins.

Yeah.

Than trying to solve the problem
that you've not really touched

previously, but only occurs
maybe once a month at the close

of the month or something.

Those are not that those aren't
worthwhile to solve.

No, but I mean, that's not where
you would normally think, right?

I mean, that's not what you
would immediately go towards.

Right.

So now I do want people to

understand that like, because it

is tempting to, to make a big

splash.

And I mean, not that there's
egos behind it, but there's

pressure behind it.

There's pressure.

And so it, you know, hey, we, we
automated this massive process.

Sounds better to a board.

Yeah.

So or to the shareholders.

But at the same time, the ROI

and AI that we talk about a lot

is really getting in those

little, like you said, those

little things that are just,

they're just, they're actually

kind of almost obvious that you

don't so obvious you don't see

him.

Exactly.

They're, they're kind of the,

the mundane task that you sort

of just gotten used to doing

every day.

And, and in fact, in, in some
ways, it sort of defines the

role to today, but there's
really an opportunity to kind of

change the role for that
particular function in a

business such that they're,
they're, they're spending less

time doing this low value work
to gather information.

Yeah.

And they can have the agent or

AI in general prepare that

information for them to make

those decisions.

So, um, we'll start wrapping
this up here in a minute.

Yeah sure.

Practical business outcomes, um,
you know, like thinking of

faster claims processing or
reduced onboarding for HR.

You guys have talked about some
of those.

What are a couple others you're
seeing people, I mean, I'm

looking at a screen running off
camera over here, and it's just

got lists of them.

Uh, there's so many options.

Yeah.

What are some of the ones that
you guys are seeing?

I know HR really is adapting
this.

We're talking, I had some

conversations this morning about

that and just so much great

opportunity there.

What are some other ones?

Um, I would say, you know,
certainly AP's been a long

history of focus for Hyland and
our ecosystem alike.

Um, but health and human

services and government

agencies.

Yep.

You know, that are all really

stressed for, for funding right

now and the opportunity to

continue to serve this kind of

growing demand of digital

interaction from citizens and

constituents.

Mhm.

Um, in a way that you don't

really have the budget to staff

for more, you know, more FTE for

that.

And, um, you know, I think that

that's, I often look at the

opportunity that exists there to

meet the growing demand on your

organization.

Yeah.

Um, with the, the, the same, you

know, investment and in terms of

systems and resources and, and

people combined, knowing that

you, you continue to need human

in the loop decision making in

that.

So I, you know, I think there's
real opportunity there in those

other industries and, and, and
solutions and verticals where

you see this growing kind of
demand of interaction and, and,

and speed to decision is
critical to the, to the process.

Yeah. I think you guys, you
know, we work together closely

and just being able to help
people see what it looks like,

you know, um, demos are great.

Yeah.

Um, but being able to apply it
to your own, your own situation

is super, super valuable.

I think one of the things we
close I want to close on is I

think there's a temptation to
have to, you know, replace your

existing systems or something.

And I don't I don't know that

that's true, but there may be a

use case where that's important,

right?

I mean, it may maybe you do need
to.

I think most organizations I
kind of started, I think said

this earlier, they already have
some sort of repository system.

What's the first step is we kind
of leave, you know, put a bow on

this first step for somebody who
maybe it's the COO, maybe it's

the CTO, maybe it's the CEO who
knows, you know, an IT leader

that's tasked with doing this.

Hey, I want to move into that
intelligence layer.

Mhm.

What, what, what are the first
steps or questions that is going

to get them in the right headed
in the right direction?

Obviously, we would sit here and

say, they need to call one of

us.

Right?

Right.

Yeah.

But but I mean, just in general
terms, what's going to help

people take that next step and
feel confident about it?

Speaker 3: I think there's a couple of key pieces.

One that we've sort of already
touched on is where are your

most attainable wins?

But then closely following that

is the data that's inside the

documents themselves.

So historically, it has been
sort of about ensuring that that

data is secured and sorted, you
know, in a digital cabinet, you

know, in an ECM platform and
that that met the need.

But now you're talking about

exposing, you know, individual

characters on the page to other

tools and understanding what

might be housed in those pages

that is sensitive or could

potentially be misleading to an

AI, AI platform that's consuming

that information.

Um, and beginning to, to really

think about that unstructured

content as a data component,

right?

And how you might need to

instruct the AI platform as it's

consuming and curating that

information, right.

And provide input to it.

Um, but then also really

understanding any sort of

compliance controls that you

might have around that things

that might need to be completely

ignored.

Um, you know, in that curated
data set that you're using to,

to fuel this.

So, so.

Little Hyland plug here
opportunity.

Right?

Um, if you're a Hyland customer
already.

Mhm.

Um, and you're not in, in the
CIC, well we call it CIC, but

the content based cloud, right?

Content cloud.

If you're not looking into that
or already have it, what, what

is the advice to customers to do
first, like reach out to their

rep, reach out.

Obviously that's it.

Ask questions.

Um, but what are they missing by
not right?

Because I think it's, I think

it's important to go, hey,

you're really missing out a lot

of intelligence.

That's great.

That's big.

That's broad.

Yes.

Intelligence.

I see the value of that.

We've moved from this to this.

I think most people are starting
to grasp, but I've got a sense

of that just being around this
year versus last year.

Yeah, I think as people are
starting to really go, instead

of kind of bringing them to the
conversation, they're kind of

coming to the conversation now,
which I think is great.

Yes.

Um, but for Hyland customers,

like, what are you most excited

about?

Where should they, where should
they get excited about?

Speaker 3: I think really.

Beginning to, to understand the
opportunity and maybe a little

bit of what you might be
alluding to there is we've been

doing a lot of discovery
workshops with customers and,

and really kind of helping them
champion it internally.

Yeah.

Um, and, and being able to sort
of, one, maybe grow their role

in the organization, you know,
as the champion of this, this

realm of possibility, right?

That's there.

But there are, you know, there

are kind of layers of, of

opportunity there.

The, layer that's probably most
easily understood by long time

ECM veterans sure is IDP.

Right, right.

You know, we've been doing
capture and we've been doing

forms of IDP for decades, right?

Um, but, you know, kind of the

advantages with that, but then

that's maybe an initial stepping

stone, right?

And I, I think the, the

excitement that I'm starting to

definitely see percolate here at

the event this year is around

the opportunities for agentic

automation capabilities sort of

embedded within existing

workflows and how they can, how

they can leverage that to, um,

further optimize, maybe identify

FOD and risk within those

workflows.

So it's not always about
efficiency.

Yeah.

You know, um, it, it might be

avoiding risk within the

process.

Or building new workflows or

building new workflows, like

super helpful.

You know, it's powerful what you
guys have shown.

Yeah.

Yeah, absolutely.

And, uh, you know, there are
advances both in sort of the,

um, in delivered solution as
well as the administration side

of the equation, um, with sort
of natural language, prompt

based solution administration
that continue to advance.

So I think that's a, that's an
interesting frontier for a lot

of the admins.

Yeah.

Whereas the business owners are
more focused on, you know,

efficiency optimizations and
outcomes in that equation.

That's a good point.

Yeah, that's, that's excellent.

And, and what's really cool is,
um, yeah, it's just I, it's so

much of is happening.

It's just so fast.

We talk about that all the time.

You know, it's happening so fast
and it really is easy to get

your head just spinning.

Yeah.

Um, because even just in your

last comment there, you talked

about, we just, we hit three or

four different things that like

if somebody was like, oh, I want

that, oh, I want that, like

you're going to kid in a candy

shop.

I want that one, that one, that

one, that one, just to take all

of it.

but we just want to.

Hey.

Slow down.

Mhm.

Ask questions like, I think
that's the most valuable thing

is when I say we're all trying
to figure it out, some of us are

further along than others.

I'm just saying it's there's
constantly evolving.

So we're all learning and
growing together.

Um, nobody's perfected anything
like completely.

Right.

But, um, I think you guys have
done just a great job of making

this agentic opportunity open to
like your entire product line.

And then we're super excited
about, about helping people out.

Thanks for joining me, man.

Glad to.

Yeah, it's always a good
conversation.

Yeah. So thanks for joining us
on an Unstructured podcast.

If we said anything brilliant,
it was probably him just as soon

we planned it.

We'll see you next time.