WWelcome back to another episode of Tech
on Unhing where tech gets human. Today
I’m joined by Patri Sheen, director of
sales engineering at Cresendo. Patri
spent over 15 years in sales engineering
from Comcast to hardray fiber and now
Cresendo, helping enterprise companies
adopt new technologies the right way.
He’s one of those people who knows how
to take big exact visions and turn them
into solutions that actually work. which
is exactly why I am super excited to
have you today with us, Patrick.
>> Thanks for having me, Ra. Yeah, thank
you.
>> You’ve spent a big part of your career
in technical sales leadership. What’s
the most surprising or funny
misunderstanding you’ve ever had with a
client about AI? Well, the funniest
thing is a lot of people will will look
at AI tools from a demo and they
extrapolate how it would, you know, work
in practical environments and
essentially they would see a tool like
um, you know, AI sentiment analysis on a
demo. And we have customers that that
would thinks that the sentiment analysis
is basically a way for an agent during
the interaction to course correct.
They’re thinking that the sentiment
analysis which is really in you know
after that interaction to maybe coach
that agent on what they did wrong, what
they did right but they are expecting
that this is real time shaping that
agent’s interaction. And so the funniest
thing we would see or the the weirdest
thing that I would see is that
expectation is that oh we we thought
this tool would make our agents in real
time change their behavior and this is
the guardrail that this is where I have
to step in from time to time and and
really hey AI tools are great that
really not magical in that sense they
they don’t change people’s personalities
uh to that level. So that’s that always
strikes me as uh what’s what’s real and
what’s expected and I think that a lot
of the the noise feeds into that right
so that’s that’s my biggest observation
and and that can be extrapolated to
other tools beyond the uh sentiment
analysis or things that like that but AI
has a tendency to make people think that
a lot of things can change in a way that
is really not you know based on reality.
Patrick, now if we break down the jargon
a bit, we hear terms like AI governance
and responsible adoption tossed around a
lot, right? What do you actually mean um
by these terms for enterprises?
>> Yeah. So when we’re talking about
governance and accountability, this is
where my department and I come in is
where there’s a lot of tools that are
being added and promoted so fast that
there’s a lot of compliance and and laws
and regulations around them have not
been fleshed out or established. So it’s
on those providers to really set the
stage and and really keep things
practical as opposed to what customer
may wish what they are being sold.
Right? So my contribution really comes
in where I have to step in and not be a
a blockage or a not slowing down the the
progress but really to be the grown-up
in a room in a sense. Ask the right
questions like okay AI can do this but
where is it getting the the data? Is it
safe? And you know the integration
portion of it does it have access to
more data than it needs? Um how is that
data contained? So we have things like u
different environments like you know
medical environments that would need to
have um some data presented to them by
AI and how does that fit with HEPA
compliance for example or how does that
fit with PCI or different compliance
based on the industry right so that’s
the part of it where essentially again
the the grown-up in the room making sure
that things are are done in a way that
keeps the trust things can be good on a
demo or things can look great it’s on us
from the technical side to really
understanding the back end and what
security features are established and
what what compliance are maintained and
so on. So,
>> so Patrick, when executives say, you
know, they want to build an AI first
company and you might get a lot of leads
on that, what is usually the simplest
way to explain the gap between, you
know, their ambition and the reality of
it?
>> That’s a great question. best way
whenever uh we have a C executive or a
leader either within our own
organizations or organizations that we
are interacting with whenever they need
some tool or they want to ask if you
know certain AI tools are are are needed
or necessary we have to analyze a few
things the first set of things that I
analyze is really what is the workflow
how does AI fit in the workflow because
AI does not reside in a vacuum
especially when we’re dealing with
systems like a SAS software as a service
and we tack on different systems to make
the whole flow. So we have to ask the
question where does AI fit within that
that flow and is there any type of
disruption by introducing that AI tool
how will that interact with the other
tools that were there right so I’ll give
you a quick example if for example they
wanted to add an AI tool on the front
end so maybe an AI receptionist or an AI
agent that will take incoming
interaction incoming calls and route
those calls right so my question would
be okay that AI How does that integrate
with the with the CRM or the system that
manages their customers information?
Right? So when a caller calls, maybe
they have a system that has that
caller’s information so that typically
that information gets presented to the
agent so they know, hey, I’m talking to
Robia and you know I’m ready to interact
with you because I have a list of the
last times you call and all of that,
right? If we’re introducing a tool that
cannot integrate with that system, then
no matter how powerful that tool is, if
it does not integrate with the with the
source of the data, it’s really not
useful in any meaningful way. Right? So,
I have to analyze that because when we
have a seale executive that comes in and
say, “Hey, I’ve seen this in a demo. I I
want this available.” They’re not seeing
that part of it that’s missing because
in a demo you have a container
environment. uh there’s no integration
that’s needed really but it’s my job to
really look at the bigger picture and
all the problems and the restraints that
we need to um establish.
>> Yeah. Yeah. Absolutely. You know Patrick
we do see a lot in the tech landscape
debates um the argument around you know
how the sea can be the biggest risk to
AI adoption. So you know from your
perspective how true is that?
>> It’s very true and and it it varies on
on environment. If you have a seuite
that’s made up of not necessarily
technical people or people who came from
a technical background that’s not
necessary but people who have the
humility or or the curiosity to really
understand what they don’t know and ask
the right people or if you have a seuite
that will include maybe some some of the
dev members or some of the engineers as
part of those conversations that is the
the best way but what I’ve seen in are
in a lot of different environments, not
my own, but I’ve seen a lot in different
environments where you have the
seuitees, they they go to conferences,
they they get captured by the AI boom.
They they’re seeing uh other companies
or their competitors investing a lot in
AI and they I want this tool right now.
No matter what, I want it now. And if
you’re a C executive and and you have
those strong requirements without you
know consulting with the dev team and
you make it as a this is what we want
make it happen. In most cases uh the dev
team can make it happen right if you ask
the dev team to hey I have I need hot
water in the shower right and the dev
team can find a way to make the the
pipes hot enough when the water comes
out it’s warm but that really doesn’t
address the problem. It’s kind of like
putting a bandage in in a problem. So
that’s what I’ve seen. I’ve seen
environments where seuitees do not
really understand what it takes to build
and maintain systems and they just want
to see the results and this is where you
know the executives themselves if they
have a technical background they can
circumvent that. they would not they
would understand what it takes or if
they communicate if they have a an open
door policy where um the tech team can
really have some criticism and can
really have their say and they really
take that in consideration then they can
avoid that but that has not been that’s
not what I’m seeing at large
>> so you know if you sort of bring it down
to a few options so where do you see
executives most often underestimating
the complexity of AI is it usually the
integration compliance or the reality of
scaling.
>> The biggest thing is really integration
and that is the biggest topic next to AI
itself because integration requires a
few things that again is really not part
of the conversation. It’s really
technical, right? But to break it down
is all those tools are not being
developed by the systems themselves. So
if you’re talking to a telecommunication
system, you’re talking about to a CRM
systems, they’re not developing those AI
tools. They are using them. um and they
might configure them to fit their
environment. But when we’re let’s say
we’re selling a system to a university
or we’re selling a system to a school or
or medical facility, they have systems
already in place that they’re using,
right? They have system already in place
that they they need to continue using.
So the integration of the AI tools or
the system that we’re selling, how does
that integrate? How does that system use
the information from the other system to
be part of that workflow and that is not
usually part of the conversation not
from the pre-sale all the way down to
until things need to be implemented
right so that’s where I see the biggest
challenge is the integration because
systems to integrate to a system it’s
not just to create that what we call a
web hook to create that connection but
once that system connects X how is that
data managed in that new system right is
that is that data managed in a way that
makes the tool easy to extrapolate or
take out the the necessary data is that
data clean because no matter how
powerful the AI tool is if you feed it
bad information or bad data you’re going
to have bad results right so that’s the
part that I see the biggest challenge
the other part is again what we talked
about earlier is really the uh guard
rails on security on compliance and all
of those things because I mentioned that
we do have some compliance in place, but
because of how fast AI tools are popping
up and how how quickly those things um
need to be implemented that the the
compliances are around AI and AI safety
are are are not there yet. They’re not
find hasn’t been enough time for those
compliances to be to be around. So it is
on company itself providing that service
to make sure that okay we need access to
some of the information when we process
that information you know what type of
guardrails we need to put in place what
type of gotchas we need to really be
aware of so those are the two aspects
that I think people should be really
focusing more on. Have you come across
situations where executive enthusiasm
pushed a company into an AI initiative
that was unrealistic or poorly planned?
Um is there a story to share? Can you
tell us what happened?
>> Uh yeah, without without mentioning
names, um we we are working right now
with a with a major uh medical uh
provider, right? And they are they are
coming off of a a system that they
purchased about a year ago. And again,
the system was was sold and they had and
this is based on on them when we brought
them on. They they said they had a great
demo. The system they they they showed
the system doing what they wanted it to
do and then they got the system
integrated. They got the system
implemented and turns out that the
system does not really uh communicate
with their EHR system. This is the
system where they have for medical
facilities. This is a system they have
where they manage their their patients
information where they match. So you can
picture a lot of HIPPA a lot of
compliance around people’s medical
information. What ends up happening is
you know 6 months down the line they
find out that their process because of
the lack of integration the process is a
bit even even worse than it was before
because now the agents or or the nurse
or whoever is using the system now has
to still rely on their old way of
interacting with the patients
information and add on to the layer of
okay now I’m dealing with sending an
email or responding to a web chat or
answering a call. So now instead of
having that one one portal, one system
that integrates with everything that
makes things easier for me. Now I have
multiple screens or now I have multiple
systems. And the more systems you have,
the slower the whole process becomes. So
they’re really in a situation now they
have a new system that doesn’t quite
integrate with the way it was promised
and now the whole workflow becomes a lot
harder for them. So what ends up
happening then is we’re we’ve been
working with them for a few months and
they’re very careful, you know, rightly
so, right? They they want to have proof
of concept of every step, right? So they
don’t fall back into uh the the
situation that happened because it’s
really in terms of ROI it return on
investment or in terms of the the
problem that were caused by adding a a
badly configured system into a major
medical environment is unimaginable.
Right? So they’re very careful and we’ve
been working with them and it’s sort of
like having you know the teacher looking
at your homework uh every step of the
way. It makes us better as as as an
implementation team. But we can tell
that okay, they’ve been burned before.
So it that’s what happens now. We have
to show the integration. We have to show
the integration at in practical
environments under stress. We have to
show what happens when the system breaks
down. How quickly can it get back up?
What are the other ways of interacting
with system if this is not available? So
there’s a lot of things that we have to
do now be to course correct or to sort
of like when somebody comes from a bad
breakup, right? And they’re meeting
somebody new, they’re a little bit more
careful. So this is what what’s been
happening. There are other uh situations
like that. We’re we’re going through the
same thing with her with her school. Uh
and that’s that’s different because they
they were promised a web chat that was
able to do certain things. And then
turns out that again another issue with
information being pushed, integration
not not uh following through, data not
being managed the right way and all of
those things. you know uh uh Patrick
when we look at the stats today um we
see that enterprise IT is you know uh
the entire spend is pushing up till$15
trillion or more right so do you think
AI adoption is sometimes more about just
optics you know the shiny object
syndrome then the real outcomes of it
>> it there’s there’s definitely an aspect
of that right with with about five you
know five trillion dollars in IT AI
spent right it’s really there’s a lot of
boast ing there’s a lot of okay my
company is is investing that much in AI
but how much of that is actually
practical it’s it’s very murky right now
because you’re betting on things that
have not had enough time to to show a
return on investment you’re betting on
things that there’s a lot of wishful
thinking on that but it’s necessary
because again you bet on an AI tool that
has the potential of really changing and
really changing the whole stack of how
certain industries are are using those
tools. So if you ask me the amount of
money and the amount of resource that is
being spent I think is necessary but I
think there are some of the these
guardrails that we spoke about earlier
that needs to be in place so that if a
company has a budget of let’s say you
know a billion dollars on AI my advice
would not to be to reduce that
investment is to really understand where
that is going and have some guardrails
around how that money is being spent and
>> so um you know given that it’s 15 years
of expertise Patrick that you’ve had in
this particular tech industry and then
in sales. So in your role as a sales
engineer, how do you balance supporting
executive vision with being honest and
what is being realistic and when do you
know you know it’s time to push back and
just you know being straight away
honest.
>> Great question. I I have a set of
questions that I ask. They’re very
poignant. They’re very there’s a reason
why I ask those questions, right? when
I’m looking at what problem is the AI
solving, right? Is there a way any other
way to solve this problem without the
complexities of integration and AI and
so on? Is there a a a simpler way to
solve this problem? So the question is
like are you asking for this tool
because it sounds good or because the
competitor has it or because you saw it
on a demo or is this something that can
actually solve an actual problem that
you have? Right? And then once I have an
answer to that then I start asking okay
how would that AI fit in the whole
workflow from beginning to end right? So
I try to understand the current
environment and then see where the AI
would fit as opposed to pushing an AI
tool and waiting for the results and
hoping that hey it it worked right. So
when I’m in in conversation with the
seauite and you know um directors or or
leaders I typically ask those questions
and understanding what the answers are.
If you ask a leader, um, give me walk me
through your process for a quick
example. Tell me what happens when I
call your hospital. Um, who do I speak
with? Right? And how is my call routed?
Who picks up that call? My information
that gets captured, how is it captured?
Is it the agent themsel that has to type
in or write down? Is it captured
automatically? Where does that data go?
And then once that data is captured, how
are you keeping that data safe? What
system do you use? And when you run your
report, do you have a need for all that
data or you just capturing data just for
the sake of capturing data? Right? And
so I go down that list and the the
longer that conversation goes or or the
the more information we capture from the
front end, the less friction we get on
the back end and the more tied to
reality we we get. Right? And if those
questions are not being asked now,
you’re at the whim of the people selling
you those tools, right? Because because
they want to sell those tools. Those
tools need to be sold, those needs to be
implemented. And you know, you worry
about those issues later on, which is
not the right way to to go. So you know
when it comes to technical sales
leadership while keeping the product
legal or business aligned to um you know
how they are contributing or what value
addition they are doing within their
tech spaces. So how do you guys handle
the AI adoption at that particular stage
or point?
>> When we’re looking at different
different departments of different
groups that are involved in in the
process they each have a different part
of that role. So I understand like there
for example the seuite they really don’t
need to be that technical but they need
to understand the the requirements and
what’s possible and what’s not right the
sales engineering team needs to educate
the sales reps the sales reps who are
promoting the systems or promoting the
AI tools they need to understand the
practical application and the
limitations of those tools so they don’t
really need to understand how the
sausage is made but they need to
understand what tools are currently
available and what those tools can and
cannot do yet, right? Because they have
to when the sales team or when you have
somebody promoting the a particular AI
tool, they have to promote it with
what’s currently available, right? And
and if it’s a tool that’s not released
yet, that is, you know, coming later on
or if it’s a tool that is in
development, in beta development, that’s
not really uh that has not been stress
tested, they have to mention that,
right? And this is the the part where
that our department when we educate the
sales team we make sure that they
mention those things right and then the
implementation team right so now you’re
you’re going from the big picture the
sea level that the executives making
those okay we can offer this now we want
this tool to the people selling it the
marketing to the people implementing it
because once you sell it you have to
implement it you have to make it work in
that environment so that team needs to
have a different set of training. So
that team needs to understand what
customer wants and how to relay that
information to the development team who
are building and configuring the system.
So that’s the implementation team and
then you have our devs. Those guys they
need as guys and girls they need to
really be uh very technical because not
only do they need to know different AI
tools out there they need to know the
integrations between those AIs and the
various system that they need to
integrate with. They need to understand
the the the limitation and the
parameters around data and data access.
So different layers you have different
knowledge or different different things
that need to be acquired for them to be
effective. So it’s it’s really important
to mention that not everyone will know
everything. So we have to separate what
every every piece of the puzzle needs to
what every department need to learn. And
this is where we come in. When I do a
training I have a specific training for
you know the C executives. I have a
specific training for the sales team, a
specific training for our implementation
team and so on.
>> No, that makes a lot of sense. And you
know since Patri you work with a lot of
international clients, how do you see
regional regulations like you know
Europe’s AI act shaping executive
decisions differently from you know the
US
>> things will be a bit more um they they
will end up converging because you have
it’s not you can’t be as siloed as you
you were not even a decade ago. So
things are more integrated. So companies
now are more and more global. There are
still some regulations that are applied
in different places and and others
different markets and so on. This will
always be the case. But when it comes to
AI and how data is managed and so on,
you will see a convergence of those and
we have not seen it yet. So
internationally I know that there are
certain parts that are ahead, right?
certain places that are had some some
places are are more loose with their
regulations. So this is these are the
places where you’re going to see more of
the problems first. It’s really a the
places where AI is being pushed, you see
AI everywhere is usually where you’re
going to see some of the issues come to
light first. There’s been not a lot of
regulations versus places where there’s
more of a slowdown in adoption, right?
and they either way you can go too far,
right? If you restrict it too much, then
you’re not having a lot of ways to have
the AI tool being stress tested in a
practical environment, right? Because in
a lab, things can work well in a lab,
right? In a lab, we’re not tied by um
regulations. we can test all those tools
but then once those tools are applied
and for us to be able to see the tool in
action in a real life scenarios this is
where regulation steps in. So the the
opportunity for a lot of those systems
to be stress tested in a real
environment is is being slowed down by a
lack of um um regulation or too many
regulations around it depending on where
you’re looking at it. I think I think
the the US market which is the biggest
market that that I work in without being
biased I think this is the sweet spot
where there are some regulations there
uh we we we need some better ones
obviously but I’m looking at the market
I’m looking at you know the Asian market
and different places I think that based
on what I’ve seen the US market has the
the right balance even even though it’s
not quite where it needs to be.
>> Yeah. Yeah. So, um you know, if we talk
a bit about the diversity and inclusion
of AI, we see that AI and tech
leadership are still often, you know,
criticized for lacking diversity.
There’s not a lot of room for um you
know, um creativity there. So, how do
you think more inclusive leadership
changes, you know, the way organizations
adopt AI responsibly? That’s great
observation and and I I remember the
best uh one of the best AI communicator
that I’ve seen was um last year at a
another Taris conference was I forgot
her name. Um she had a a real a real
good human way of promoting AI,
promoting AI tools and the way I’ve
never seen somebody communicate AI that
way. And then what you’re seeing here
and and and we definitely need more of
that. So what you’re seeing here is you
have in a tech world you have which is
dominated by by you know male engineers
and so on you have those I would say
those gatekeeping processes where to
talk about AI or to promote AI you need
to be very technical which is which is
really not not needed and at some point
it could even be a a bad thing if you’re
only seeing things from a technical lens
even if you’re talking about something
highly technical as as AI. So I think
there should be a place for you know a
lot more diversity, a lot more different
ways of communicating AI and people
relate to people that can talk to them
in a way that that the communication or
the the information can be relayed and
can and I think that that’s what’s
lacking from a C level all the way down
to the development teams who are
actively using those those tools. And
from from what I’ve seen, every time we
have more integration, every time we
have more women, more people of color,
more people with different voices, as
long as they can understand and and and
see the big picture, I think I think we
need more of that. Again, I’ve I’ve seen
great communicators and what I’ve
captured from them as they understand
the system. They’re not trying to show
that they’re very technical, but they
just want to explain in real terms in in
day-to-day how those tools can be
helpful. And I uh agree with your
assessment here. I I’ve not seen enough
diverse representation in that space.
It’s a continuation of the tech space.
>> Yeah, absolutely. Another blind spot
that I’d like you to talk a bit about on
is the vendor lock in versus
flexibility. And that would you know be
more educational for our listeners as
well you know when they look out for um
you know expansions. So we see that how
many enterprises rush into AI vendor
partnerships without thinking long term.
We are operating on the short-term you
know um instincts or just want to hop
onto the bandwagon of it. So you know do
you see vendor lock in as another blind
spot executives are missing?
>> Yes. Yes. That’s that’s another big big
um issue there is uh people are being
locked into AI tools that you know are
not um because AI is changing a lot fast
and sometimes it’s not you can have an
AI tool that will do a function and a
newer version of of that AI comes out 6
months later. It’s not in many cases
it’s not as simple as re removing this
one and putting the new one. It’s not
like you’re getting a a new tire and
just change the tire. If you have a
system that is hard to integrate and in
order to have this AI tool that just
released to work for your system,
there’s a lot of back-end patchwork that
needed to happen. And now, you know, our
dev team or your dev team made it work.
Okay? And then 6 months later, there’s a
new version of that AI tool that will
do, you know, more than than what the
previous version did. Okay? Now you’re
locked in into a contract and there’s no
way of, you know, removing the old tool
and adding the new tool. And I’ve even
seen environments where they had the old
tool, they still need the new tool. So
they add the new tool on top of the old
tool. So now you have two you have an AI
autoattendant, right? When you call in
the company, you have an AI system that
can route the call, which is which is an
older version. And on top of that,
they’re adding an AI agent that can
actually take the call and and then
process what the customer is asking and
then route it and so on so forth. So now
because they they’re locked in using the
old AI tool, they’re still using the old
AI tool on top of the new one that’s
added and that creates all kinds of
issues that are on the back end makes
the system really not not that great
when you look under the hood. Yeah. And
and so so the best system are systems
that are built for integration that you
can you can take things out, put things
back in and whatever you’re putting in
to make sure that that tool can be um
updated, that tool can easily be, you
know, you go from version one to version
two and from version two and on. So
that’s what I think people should should
really be focused on. Which is why uh
when we talk about AI, I always want to
bring it back to integration, right?
Integration is really what makes AI what
it is until we get to AGI.
>> So Patrick, you’ve built your career at
the intersection of sales and tech.
What’s one leadership lesson you’ve
learned about bridging vision with
execution? Yeah, I’m in a I’m in a very
um I wouldn’t say unique uh place, but I
I was in technical sales while I was
going to school for engineering. So So
while I was getting my technical
knowledge, I was actively selling and
then those are two very different
mindsets, two very different set of
skills, right, that sometimes conflict
with each other. So I I was um fortunate
enough to to build my sales career
before I got indoctrinated into the tech
world. The biggest challenge is really
understanding tech, but being able to
talk to people who have no desire to to
know tech, right? But they know they
they need it, right? So throughout my my
career, I’ve I’ve developed that that
skill where I understand it and I can
geek out with the the devs and then uh
when I need to explain tech to
executives or or sales leaders who just
want this to work, there’s a translation
that that that is not is not as as
simple as it may seem, right? So I’ve
had to learn how to do that. And the
biggest lesson that I’ve learned is and
and really that ties into the technical
mindset is I’ve seen sales engineers or
or developers talking to executives and
and the way that conversation flows is
is really they’re talking above their
heads and thinking that what they’re
saying makes sense because in in their
mind it it does. But it it it’s really
not lending because you’re talking to
somebody who has a big picture and
you’re talking about the fine the the
details, right? Being able to parse out
the detail, although important, but when
you’re when you’re talking to to people
who are not technical, there’s a way to
communicate that is that is, you know,
uh very different than how text need to
communicate with each other and so on.
So I think that that’s the biggest
lesson that I learned and I I did not
wake up one day and decide hey I’m going
to need to to learn this. It’s just over
year over the years I’ve I’ve realized
that okay I’m selling a technical
product. The more I know about the
product the more opportunities I can
find but at the same time I don’t want
to go in and say you know and that was
before the talk about AI we’re talking
about systems back end we were talking
about integration was still at play.
We’re talking about uh automation. We’re
talking about uh things like now we’re
talking about uh web chat that is
powered by AI. We 10 years ago we’re
talking about web chats, you know, uh
rules-based web chat where they mimic
what we now understand as AI, but they
they were rules-based, right? Press u
you know, press this to go there. So, it
was like a mapping structure in a back
end that simulated. I knew that when
you’re communicating with people who who
are not interested in in sales really uh
in in in tech really there’s a way to
communicate with them. That’s the
biggest lesson I
>> learn. Yeah. Yeah. No, I think that was
you know that was a great insight that
you um shared but also um Patrick you
know um extending you know the
conversation that you were doing
earlier. Fast forward 5 years what will
we look back on as the biggest mistake
enterprises made with the AI adoption?
five years and it could be it could be
because because the way AI is is going
it could we could say five years now and
then a year later our prediction as it
relates to AI has always been uh uh we
predicted something would happen in 10
years and it happened in five every time
we make a prediction as it relates to AI
things happen faster than than than we
anticipated so let’s say five years from
now the big the mistake that a lot of
organizations are making is one they
have not done the prep work for getting
their database AI ready. Two, they’ve
not gotten their development team, their
implementation team trained on how to
integrate the AI tools that are coming
out faster, smarter with their existing
systems. Third probably would be being
locked into AI tools that that were not
built with integration in mind. AI tools
that were just packed on to on top of,
you know, systems that you can you
cannot remove. They’re not listening to
their to their sales engineers. Really,
that’s that’s that’s what it boils down
to. Uh they’re not listening to the to
to the team that’s in the middle between
reality and wishful thinking. I think
looking back, that’s where the mistakes
would be made.
>> Yeah. Well, we’ve recorded you and any
year after we will, you know, redo this.
um entire podcast featuring you and
telling the world how Patri already
predicted that
>> how I predicted how how my prediction
came to to pass and and don’t don’t
shoot the messenger.
>> Yeah, absolutely.
So, um Patrick, moving towards the last
question of our conversation. If you
could give one bold piece of advice to
today’s executives on how not to derail
adoption, what would it be? All right.
And if I could really summarize it for
the executives, it it would be this.
It’s very simple to look at a tool uh
whether it’s a it’s a conference or or a
commercial or somewhere to look at a
tool and seeing the vision. The best
advice I would give um to executives
would be to um to really listen to their
internal team, listen to the people who
are implementing those tools. If we’re
if I’m talking to to executives who are
selling or or implementing this into
their system to sell and if I were
talking to executives who are who are
looking to acquire AI tools for their
for their system to re to really ask
themselves these questions. One is what
specific problem am I trying to solve
and does this tool uh uniquely can solve
this problem? So is there any other way
for me to solve is is this tool the the
best way for me to solve this problem.
So that’s the first part. The second
part is how does this this tool fit in
my current environment because when you
acquire AI it you’re not acquiring a
standalone system. You’re acquiring a
system that will need to live within an
ecosystem. So, do I get that AI and then
worry about my ecosystem and and and how
it integrates later or do I make sure
that when I get the AI, I make sure that
it can integrate with the system that I
value that I need or best way is even
before you get your AI to make sure that
your system is ready for integration.
Your system your your data is clean,
your process is you know laid out,
right? what do people need to do from
point A to Z and then you bring in that
AI or you can partner with uh companies
who are uh promoting and pushing AIs.
You want to partner with with those
companies who will, you know, point
those things out to you. Not just sell
you the tool, but really asking, hey,
let’s take a look at your system right
now. Let’s take a look at your flow.
Maybe we need to do that. Maybe we can
revisit this in 6 months. In the
meantime, this is what you want to do.
That would be my my advice to them.
>> Well, Patrick, this has been a valuable
conversation. Thank you so much for your
time and insights.
>> Thank you so much, Ravia. It’s been a
pleasure.
elcome back to another episode of Tech
on Unhing where tech gets human. Today
I’m joined by Patri Sheen, director of
sales engineering at Cresendo. Patri
spent over 15 years in sales engineering
from Comcast to hardray fiber and now
Cresendo, helping enterprise companies
adopt new technologies the right way.
He’s one of those people who knows how
to take big exact visions and turn them
into solutions that actually work. which
is exactly why I am super excited to
have you today with us, Patrick.
>> Thanks for having me, Ra. Yeah, thank
you.
>> You’ve spent a big part of your career
in technical sales leadership. What’s
the most surprising or funny
misunderstanding you’ve ever had with a
client about AI? Well, the funniest
thing is a lot of people will will look
at AI tools from a demo and they
extrapolate how it would, you know, work
in practical environments and
essentially they would see a tool like
um, you know, AI sentiment analysis on a
demo. And we have customers that that
would thinks that the sentiment analysis
is basically a way for an agent during
the interaction to course correct.
They’re thinking that the sentiment
analysis which is really in you know
after that interaction to maybe coach
that agent on what they did wrong, what
they did right but they are expecting
that this is real time shaping that
agent’s interaction. And so the funniest
thing we would see or the the weirdest
thing that I would see is that
expectation is that oh we we thought
this tool would make our agents in real
time change their behavior and this is
the guardrail that this is where I have
to step in from time to time and and
really hey AI tools are great that
really not magical in that sense they
they don’t change people’s personalities
uh to that level. So that’s that always
strikes me as uh what’s what’s real and
what’s expected and I think that a lot
of the the noise feeds into that right
so that’s that’s my biggest observation
and and that can be extrapolated to
other tools beyond the uh sentiment
analysis or things that like that but AI
has a tendency to make people think that
a lot of things can change in a way that
is really not you know based on reality.
Patrick, now if we break down the jargon
a bit, we hear terms like AI governance
and responsible adoption tossed around a
lot, right? What do you actually mean um
by these terms for enterprises?
>> Yeah. So when we’re talking about
governance and accountability, this is
where my department and I come in is
where there’s a lot of tools that are
being added and promoted so fast that
there’s a lot of compliance and and laws
and regulations around them have not
been fleshed out or established. So it’s
on those providers to really set the
stage and and really keep things
practical as opposed to what customer
may wish what they are being sold.
Right? So my contribution really comes
in where I have to step in and not be a
a blockage or a not slowing down the the
progress but really to be the grown-up
in a room in a sense. Ask the right
questions like okay AI can do this but
where is it getting the the data? Is it
safe? And you know the integration
portion of it does it have access to
more data than it needs? Um how is that
data contained? So we have things like u
different environments like you know
medical environments that would need to
have um some data presented to them by
AI and how does that fit with HEPA
compliance for example or how does that
fit with PCI or different compliance
based on the industry right so that’s
the part of it where essentially again
the the grown-up in the room making sure
that things are are done in a way that
keeps the trust things can be good on a
demo or things can look great it’s on us
from the technical side to really
understanding the back end and what
security features are established and
what what compliance are maintained and
so on. So,
>> so Patrick, when executives say, you
know, they want to build an AI first
company and you might get a lot of leads
on that, what is usually the simplest
way to explain the gap between, you
know, their ambition and the reality of
it?
>> That’s a great question. best way
whenever uh we have a C executive or a
leader either within our own
organizations or organizations that we
are interacting with whenever they need
some tool or they want to ask if you
know certain AI tools are are are needed
or necessary we have to analyze a few
things the first set of things that I
analyze is really what is the workflow
how does AI fit in the workflow because
AI does not reside in a vacuum
especially when we’re dealing with
systems like a SAS software as a service
and we tack on different systems to make
the whole flow. So we have to ask the
question where does AI fit within that
that flow and is there any type of
disruption by introducing that AI tool
how will that interact with the other
tools that were there right so I’ll give
you a quick example if for example they
wanted to add an AI tool on the front
end so maybe an AI receptionist or an AI
agent that will take incoming
interaction incoming calls and route
those calls right so my question would
be okay that AI How does that integrate
with the with the CRM or the system that
manages their customers information?
Right? So when a caller calls, maybe
they have a system that has that
caller’s information so that typically
that information gets presented to the
agent so they know, hey, I’m talking to
Robia and you know I’m ready to interact
with you because I have a list of the
last times you call and all of that,
right? If we’re introducing a tool that
cannot integrate with that system, then
no matter how powerful that tool is, if
it does not integrate with the with the
source of the data, it’s really not
useful in any meaningful way. Right? So,
I have to analyze that because when we
have a seale executive that comes in and
say, “Hey, I’ve seen this in a demo. I I
want this available.” They’re not seeing
that part of it that’s missing because
in a demo you have a container
environment. uh there’s no integration
that’s needed really but it’s my job to
really look at the bigger picture and
all the problems and the restraints that
we need to um establish.
>> Yeah. Yeah. Absolutely. You know Patrick
we do see a lot in the tech landscape
debates um the argument around you know
how the sea can be the biggest risk to
AI adoption. So you know from your
perspective how true is that?
>> It’s very true and and it it varies on
on environment. If you have a seuite
that’s made up of not necessarily
technical people or people who came from
a technical background that’s not
necessary but people who have the
humility or or the curiosity to really
understand what they don’t know and ask
the right people or if you have a seuite
that will include maybe some some of the
dev members or some of the engineers as
part of those conversations that is the
the best way but what I’ve seen in are
in a lot of different environments, not
my own, but I’ve seen a lot in different
environments where you have the
seuitees, they they go to conferences,
they they get captured by the AI boom.
They they’re seeing uh other companies
or their competitors investing a lot in
AI and they I want this tool right now.
No matter what, I want it now. And if
you’re a C executive and and you have
those strong requirements without you
know consulting with the dev team and
you make it as a this is what we want
make it happen. In most cases uh the dev
team can make it happen right if you ask
the dev team to hey I have I need hot
water in the shower right and the dev
team can find a way to make the the
pipes hot enough when the water comes
out it’s warm but that really doesn’t
address the problem. It’s kind of like
putting a bandage in in a problem. So
that’s what I’ve seen. I’ve seen
environments where seuitees do not
really understand what it takes to build
and maintain systems and they just want
to see the results and this is where you
know the executives themselves if they
have a technical background they can
circumvent that. they would not they
would understand what it takes or if
they communicate if they have a an open
door policy where um the tech team can
really have some criticism and can
really have their say and they really
take that in consideration then they can
avoid that but that has not been that’s
not what I’m seeing at large
>> so you know if you sort of bring it down
to a few options so where do you see
executives most often underestimating
the complexity of AI is it usually the
integration compliance or the reality of
scaling.
>> The biggest thing is really integration
and that is the biggest topic next to AI
itself because integration requires a
few things that again is really not part
of the conversation. It’s really
technical, right? But to break it down
is all those tools are not being
developed by the systems themselves. So
if you’re talking to a telecommunication
system, you’re talking about to a CRM
systems, they’re not developing those AI
tools. They are using them. um and they
might configure them to fit their
environment. But when we’re let’s say
we’re selling a system to a university
or we’re selling a system to a school or
or medical facility, they have systems
already in place that they’re using,
right? They have system already in place
that they they need to continue using.
So the integration of the AI tools or
the system that we’re selling, how does
that integrate? How does that system use
the information from the other system to
be part of that workflow and that is not
usually part of the conversation not
from the pre-sale all the way down to
until things need to be implemented
right so that’s where I see the biggest
challenge is the integration because
systems to integrate to a system it’s
not just to create that what we call a
web hook to create that connection but
once that system connects X how is that
data managed in that new system right is
that is that data managed in a way that
makes the tool easy to extrapolate or
take out the the necessary data is that
data clean because no matter how
powerful the AI tool is if you feed it
bad information or bad data you’re going
to have bad results right so that’s the
part that I see the biggest challenge
the other part is again what we talked
about earlier is really the uh guard
rails on security on compliance and all
of those things because I mentioned that
we do have some compliance in place, but
because of how fast AI tools are popping
up and how how quickly those things um
need to be implemented that the the
compliances are around AI and AI safety
are are are not there yet. They’re not
find hasn’t been enough time for those
compliances to be to be around. So it is
on company itself providing that service
to make sure that okay we need access to
some of the information when we process
that information you know what type of
guardrails we need to put in place what
type of gotchas we need to really be
aware of so those are the two aspects
that I think people should be really
focusing more on. Have you come across
situations where executive enthusiasm
pushed a company into an AI initiative
that was unrealistic or poorly planned?
Um is there a story to share? Can you
tell us what happened?
>> Uh yeah, without without mentioning
names, um we we are working right now
with a with a major uh medical uh
provider, right? And they are they are
coming off of a a system that they
purchased about a year ago. And again,
the system was was sold and they had and
this is based on on them when we brought
them on. They they said they had a great
demo. The system they they they showed
the system doing what they wanted it to
do and then they got the system
integrated. They got the system
implemented and turns out that the
system does not really uh communicate
with their EHR system. This is the
system where they have for medical
facilities. This is a system they have
where they manage their their patients
information where they match. So you can
picture a lot of HIPPA a lot of
compliance around people’s medical
information. What ends up happening is
you know 6 months down the line they
find out that their process because of
the lack of integration the process is a
bit even even worse than it was before
because now the agents or or the nurse
or whoever is using the system now has
to still rely on their old way of
interacting with the patients
information and add on to the layer of
okay now I’m dealing with sending an
email or responding to a web chat or
answering a call. So now instead of
having that one one portal, one system
that integrates with everything that
makes things easier for me. Now I have
multiple screens or now I have multiple
systems. And the more systems you have,
the slower the whole process becomes. So
they’re really in a situation now they
have a new system that doesn’t quite
integrate with the way it was promised
and now the whole workflow becomes a lot
harder for them. So what ends up
happening then is we’re we’ve been
working with them for a few months and
they’re very careful, you know, rightly
so, right? They they want to have proof
of concept of every step, right? So they
don’t fall back into uh the the
situation that happened because it’s
really in terms of ROI it return on
investment or in terms of the the
problem that were caused by adding a a
badly configured system into a major
medical environment is unimaginable.
Right? So they’re very careful and we’ve
been working with them and it’s sort of
like having you know the teacher looking
at your homework uh every step of the
way. It makes us better as as as an
implementation team. But we can tell
that okay, they’ve been burned before.
So it that’s what happens now. We have
to show the integration. We have to show
the integration at in practical
environments under stress. We have to
show what happens when the system breaks
down. How quickly can it get back up?
What are the other ways of interacting
with system if this is not available? So
there’s a lot of things that we have to
do now be to course correct or to sort
of like when somebody comes from a bad
breakup, right? And they’re meeting
somebody new, they’re a little bit more
careful. So this is what what’s been
happening. There are other uh situations
like that. We’re we’re going through the
same thing with her with her school. Uh
and that’s that’s different because they
they were promised a web chat that was
able to do certain things. And then
turns out that again another issue with
information being pushed, integration
not not uh following through, data not
being managed the right way and all of
those things. you know uh uh Patrick
when we look at the stats today um we
see that enterprise IT is you know uh
the entire spend is pushing up till$15
trillion or more right so do you think
AI adoption is sometimes more about just
optics you know the shiny object
syndrome then the real outcomes of it
>> it there’s there’s definitely an aspect
of that right with with about five you
know five trillion dollars in IT AI
spent right it’s really there’s a lot of
boast ing there’s a lot of okay my
company is is investing that much in AI
but how much of that is actually
practical it’s it’s very murky right now
because you’re betting on things that
have not had enough time to to show a
return on investment you’re betting on
things that there’s a lot of wishful
thinking on that but it’s necessary
because again you bet on an AI tool that
has the potential of really changing and
really changing the whole stack of how
certain industries are are using those
tools. So if you ask me the amount of
money and the amount of resource that is
being spent I think is necessary but I
think there are some of the these
guardrails that we spoke about earlier
that needs to be in place so that if a
company has a budget of let’s say you
know a billion dollars on AI my advice
would not to be to reduce that
investment is to really understand where
that is going and have some guardrails
around how that money is being spent and
>> so um you know given that it’s 15 years
of expertise Patrick that you’ve had in
this particular tech industry and then
in sales. So in your role as a sales
engineer, how do you balance supporting
executive vision with being honest and
what is being realistic and when do you
know you know it’s time to push back and
just you know being straight away
honest.
>> Great question. I I have a set of
questions that I ask. They’re very
poignant. They’re very there’s a reason
why I ask those questions, right? when
I’m looking at what problem is the AI
solving, right? Is there a way any other
way to solve this problem without the
complexities of integration and AI and
so on? Is there a a a simpler way to
solve this problem? So the question is
like are you asking for this tool
because it sounds good or because the
competitor has it or because you saw it
on a demo or is this something that can
actually solve an actual problem that
you have? Right? And then once I have an
answer to that then I start asking okay
how would that AI fit in the whole
workflow from beginning to end right? So
I try to understand the current
environment and then see where the AI
would fit as opposed to pushing an AI
tool and waiting for the results and
hoping that hey it it worked right. So
when I’m in in conversation with the
seauite and you know um directors or or
leaders I typically ask those questions
and understanding what the answers are.
If you ask a leader, um, give me walk me
through your process for a quick
example. Tell me what happens when I
call your hospital. Um, who do I speak
with? Right? And how is my call routed?
Who picks up that call? My information
that gets captured, how is it captured?
Is it the agent themsel that has to type
in or write down? Is it captured
automatically? Where does that data go?
And then once that data is captured, how
are you keeping that data safe? What
system do you use? And when you run your
report, do you have a need for all that
data or you just capturing data just for
the sake of capturing data? Right? And
so I go down that list and the the
longer that conversation goes or or the
the more information we capture from the
front end, the less friction we get on
the back end and the more tied to
reality we we get. Right? And if those
questions are not being asked now,
you’re at the whim of the people selling
you those tools, right? Because because
they want to sell those tools. Those
tools need to be sold, those needs to be
implemented. And you know, you worry
about those issues later on, which is
not the right way to to go. So you know
when it comes to technical sales
leadership while keeping the product
legal or business aligned to um you know
how they are contributing or what value
addition they are doing within their
tech spaces. So how do you guys handle
the AI adoption at that particular stage
or point?
>> When we’re looking at different
different departments of different
groups that are involved in in the
process they each have a different part
of that role. So I understand like there
for example the seuite they really don’t
need to be that technical but they need
to understand the the requirements and
what’s possible and what’s not right the
sales engineering team needs to educate
the sales reps the sales reps who are
promoting the systems or promoting the
AI tools they need to understand the
practical application and the
limitations of those tools so they don’t
really need to understand how the
sausage is made but they need to
understand what tools are currently
available and what those tools can and
cannot do yet, right? Because they have
to when the sales team or when you have
somebody promoting the a particular AI
tool, they have to promote it with
what’s currently available, right? And
and if it’s a tool that’s not released
yet, that is, you know, coming later on
or if it’s a tool that is in
development, in beta development, that’s
not really uh that has not been stress
tested, they have to mention that,
right? And this is the the part where
that our department when we educate the
sales team we make sure that they
mention those things right and then the
implementation team right so now you’re
you’re going from the big picture the
sea level that the executives making
those okay we can offer this now we want
this tool to the people selling it the
marketing to the people implementing it
because once you sell it you have to
implement it you have to make it work in
that environment so that team needs to
have a different set of training. So
that team needs to understand what
customer wants and how to relay that
information to the development team who
are building and configuring the system.
So that’s the implementation team and
then you have our devs. Those guys they
need as guys and girls they need to
really be uh very technical because not
only do they need to know different AI
tools out there they need to know the
integrations between those AIs and the
various system that they need to
integrate with. They need to understand
the the the limitation and the
parameters around data and data access.
So different layers you have different
knowledge or different different things
that need to be acquired for them to be
effective. So it’s it’s really important
to mention that not everyone will know
everything. So we have to separate what
every every piece of the puzzle needs to
what every department need to learn. And
this is where we come in. When I do a
training I have a specific training for
you know the C executives. I have a
specific training for the sales team, a
specific training for our implementation
team and so on.
>> No, that makes a lot of sense. And you
know since Patri you work with a lot of
international clients, how do you see
regional regulations like you know
Europe’s AI act shaping executive
decisions differently from you know the
US
>> things will be a bit more um they they
will end up converging because you have
it’s not you can’t be as siloed as you
you were not even a decade ago. So
things are more integrated. So companies
now are more and more global. There are
still some regulations that are applied
in different places and and others
different markets and so on. This will
always be the case. But when it comes to
AI and how data is managed and so on,
you will see a convergence of those and
we have not seen it yet. So
internationally I know that there are
certain parts that are ahead, right?
certain places that are had some some
places are are more loose with their
regulations. So this is these are the
places where you’re going to see more of
the problems first. It’s really a the
places where AI is being pushed, you see
AI everywhere is usually where you’re
going to see some of the issues come to
light first. There’s been not a lot of
regulations versus places where there’s
more of a slowdown in adoption, right?
and they either way you can go too far,
right? If you restrict it too much, then
you’re not having a lot of ways to have
the AI tool being stress tested in a
practical environment, right? Because in
a lab, things can work well in a lab,
right? In a lab, we’re not tied by um
regulations. we can test all those tools
but then once those tools are applied
and for us to be able to see the tool in
action in a real life scenarios this is
where regulation steps in. So the the
opportunity for a lot of those systems
to be stress tested in a real
environment is is being slowed down by a
lack of um um regulation or too many
regulations around it depending on where
you’re looking at it. I think I think
the the US market which is the biggest
market that that I work in without being
biased I think this is the sweet spot
where there are some regulations there
uh we we we need some better ones
obviously but I’m looking at the market
I’m looking at you know the Asian market
and different places I think that based
on what I’ve seen the US market has the
the right balance even even though it’s
not quite where it needs to be.
>> Yeah. Yeah. So, um you know, if we talk
a bit about the diversity and inclusion
of AI, we see that AI and tech
leadership are still often, you know,
criticized for lacking diversity.
There’s not a lot of room for um you
know, um creativity there. So, how do
you think more inclusive leadership
changes, you know, the way organizations
adopt AI responsibly? That’s great
observation and and I I remember the
best uh one of the best AI communicator
that I’ve seen was um last year at a
another Taris conference was I forgot
her name. Um she had a a real a real
good human way of promoting AI,
promoting AI tools and the way I’ve
never seen somebody communicate AI that
way. And then what you’re seeing here
and and and we definitely need more of
that. So what you’re seeing here is you
have in a tech world you have which is
dominated by by you know male engineers
and so on you have those I would say
those gatekeeping processes where to
talk about AI or to promote AI you need
to be very technical which is which is
really not not needed and at some point
it could even be a a bad thing if you’re
only seeing things from a technical lens
even if you’re talking about something
highly technical as as AI. So I think
there should be a place for you know a
lot more diversity, a lot more different
ways of communicating AI and people
relate to people that can talk to them
in a way that that the communication or
the the information can be relayed and
can and I think that that’s what’s
lacking from a C level all the way down
to the development teams who are
actively using those those tools. And
from from what I’ve seen, every time we
have more integration, every time we
have more women, more people of color,
more people with different voices, as
long as they can understand and and and
see the big picture, I think I think we
need more of that. Again, I’ve I’ve seen
great communicators and what I’ve
captured from them as they understand
the system. They’re not trying to show
that they’re very technical, but they
just want to explain in real terms in in
day-to-day how those tools can be
helpful. And I uh agree with your
assessment here. I I’ve not seen enough
diverse representation in that space.
It’s a continuation of the tech space.
>> Yeah, absolutely. Another blind spot
that I’d like you to talk a bit about on
is the vendor lock in versus
flexibility. And that would you know be
more educational for our listeners as
well you know when they look out for um
you know expansions. So we see that how
many enterprises rush into AI vendor
partnerships without thinking long term.
We are operating on the short-term you
know um instincts or just want to hop
onto the bandwagon of it. So you know do
you see vendor lock in as another blind
spot executives are missing?
>> Yes. Yes. That’s that’s another big big
um issue there is uh people are being
locked into AI tools that you know are
not um because AI is changing a lot fast
and sometimes it’s not you can have an
AI tool that will do a function and a
newer version of of that AI comes out 6
months later. It’s not in many cases
it’s not as simple as re removing this
one and putting the new one. It’s not
like you’re getting a a new tire and
just change the tire. If you have a
system that is hard to integrate and in
order to have this AI tool that just
released to work for your system,
there’s a lot of back-end patchwork that
needed to happen. And now, you know, our
dev team or your dev team made it work.
Okay? And then 6 months later, there’s a
new version of that AI tool that will
do, you know, more than than what the
previous version did. Okay? Now you’re
locked in into a contract and there’s no
way of, you know, removing the old tool
and adding the new tool. And I’ve even
seen environments where they had the old
tool, they still need the new tool. So
they add the new tool on top of the old
tool. So now you have two you have an AI
autoattendant, right? When you call in
the company, you have an AI system that
can route the call, which is which is an
older version. And on top of that,
they’re adding an AI agent that can
actually take the call and and then
process what the customer is asking and
then route it and so on so forth. So now
because they they’re locked in using the
old AI tool, they’re still using the old
AI tool on top of the new one that’s
added and that creates all kinds of
issues that are on the back end makes
the system really not not that great
when you look under the hood. Yeah. And
and so so the best system are systems
that are built for integration that you
can you can take things out, put things
back in and whatever you’re putting in
to make sure that that tool can be um
updated, that tool can easily be, you
know, you go from version one to version
two and from version two and on. So
that’s what I think people should should
really be focused on. Which is why uh
when we talk about AI, I always want to
bring it back to integration, right?
Integration is really what makes AI what
it is until we get to AGI.
>> So Patrick, you’ve built your career at
the intersection of sales and tech.
What’s one leadership lesson you’ve
learned about bridging vision with
execution? Yeah, I’m in a I’m in a very
um I wouldn’t say unique uh place, but I
I was in technical sales while I was
going to school for engineering. So So
while I was getting my technical
knowledge, I was actively selling and
then those are two very different
mindsets, two very different set of
skills, right, that sometimes conflict
with each other. So I I was um fortunate
enough to to build my sales career
before I got indoctrinated into the tech
world. The biggest challenge is really
understanding tech, but being able to
talk to people who have no desire to to
know tech, right? But they know they
they need it, right? So throughout my my
career, I’ve I’ve developed that that
skill where I understand it and I can
geek out with the the devs and then uh
when I need to explain tech to
executives or or sales leaders who just
want this to work, there’s a translation
that that that is not is not as as
simple as it may seem, right? So I’ve
had to learn how to do that. And the
biggest lesson that I’ve learned is and
and really that ties into the technical
mindset is I’ve seen sales engineers or
or developers talking to executives and
and the way that conversation flows is
is really they’re talking above their
heads and thinking that what they’re
saying makes sense because in in their
mind it it does. But it it it’s really
not lending because you’re talking to
somebody who has a big picture and
you’re talking about the fine the the
details, right? Being able to parse out
the detail, although important, but when
you’re when you’re talking to to people
who are not technical, there’s a way to
communicate that is that is, you know,
uh very different than how text need to
communicate with each other and so on.
So I think that that’s the biggest
lesson that I learned and I I did not
wake up one day and decide hey I’m going
to need to to learn this. It’s just over
year over the years I’ve I’ve realized
that okay I’m selling a technical
product. The more I know about the
product the more opportunities I can
find but at the same time I don’t want
to go in and say you know and that was
before the talk about AI we’re talking
about systems back end we were talking
about integration was still at play.
We’re talking about uh automation. We’re
talking about uh things like now we’re
talking about uh web chat that is
powered by AI. We 10 years ago we’re
talking about web chats, you know, uh
rules-based web chat where they mimic
what we now understand as AI, but they
they were rules-based, right? Press u
you know, press this to go there. So, it
was like a mapping structure in a back
end that simulated. I knew that when
you’re communicating with people who who
are not interested in in sales really uh
in in in tech really there’s a way to
communicate with them. That’s the
biggest lesson I
>> learn. Yeah. Yeah. No, I think that was
you know that was a great insight that
you um shared but also um Patrick you
know um extending you know the
conversation that you were doing
earlier. Fast forward 5 years what will
we look back on as the biggest mistake
enterprises made with the AI adoption?
five years and it could be it could be
because because the way AI is is going
it could we could say five years now and
then a year later our prediction as it
relates to AI has always been uh uh we
predicted something would happen in 10
years and it happened in five every time
we make a prediction as it relates to AI
things happen faster than than than we
anticipated so let’s say five years from
now the big the mistake that a lot of
organizations are making is one they
have not done the prep work for getting
their database AI ready. Two, they’ve
not gotten their development team, their
implementation team trained on how to
integrate the AI tools that are coming
out faster, smarter with their existing
systems. Third probably would be being
locked into AI tools that that were not
built with integration in mind. AI tools
that were just packed on to on top of,
you know, systems that you can you
cannot remove. They’re not listening to
their to their sales engineers. Really,
that’s that’s that’s what it boils down
to. Uh they’re not listening to the to
to the team that’s in the middle between
reality and wishful thinking. I think
looking back, that’s where the mistakes
would be made.
>> Yeah. Well, we’ve recorded you and any
year after we will, you know, redo this.
um entire podcast featuring you and
telling the world how Patri already
predicted that
>> how I predicted how how my prediction
came to to pass and and don’t don’t
shoot the messenger.
>> Yeah, absolutely.
So, um Patrick, moving towards the last
question of our conversation. If you
could give one bold piece of advice to
today’s executives on how not to derail
adoption, what would it be? All right.
And if I could really summarize it for
the executives, it it would be this.
It’s very simple to look at a tool uh
whether it’s a it’s a conference or or a
commercial or somewhere to look at a
tool and seeing the vision. The best
advice I would give um to executives
would be to um to really listen to their
internal team, listen to the people who
are implementing those tools. If we’re
if I’m talking to to executives who are
selling or or implementing this into
their system to sell and if I were
talking to executives who are who are
looking to acquire AI tools for their
for their system to re to really ask
themselves these questions. One is what
specific problem am I trying to solve
and does this tool uh uniquely can solve
this problem? So is there any other way
for me to solve is is this tool the the
best way for me to solve this problem.
So that’s the first part. The second
part is how does this this tool fit in
my current environment because when you
acquire AI it you’re not acquiring a
standalone system. You’re acquiring a
system that will need to live within an
ecosystem. So, do I get that AI and then
worry about my ecosystem and and and how
it integrates later or do I make sure
that when I get the AI, I make sure that
it can integrate with the system that I
value that I need or best way is even
before you get your AI to make sure that
your system is ready for integration.
Your system your your data is clean,
your process is you know laid out,
right? what do people need to do from
point A to Z and then you bring in that
AI or you can partner with uh companies
who are uh promoting and pushing AIs.
You want to partner with with those
companies who will, you know, point
those things out to you. Not just sell
you the tool, but really asking, hey,
let’s take a look at your system right
now. Let’s take a look at your flow.
Maybe we need to do that. Maybe we can
revisit this in 6 months. In the
meantime, this is what you want to do.
That would be my my advice to them.
>> Well, Patrick, this has been a valuable
conversation. Thank you so much for your
time and insights.
>> Thank you so much, Ravia. It’s been a
pleasure.