Transcript
00:00
Play from here
Copy citation
Hey, welcome to AI at Work.
00:14
Play from here
Copy citation
I'm Kevin Williams.
00:15
Play from here
Copy citation
And today we have Matt Graham, who is the CEO of RapidDevelopers.com.
00:23
Play from here
Copy citation
And based out of Boston, who's joining us, Matt has a really amazing background.
00:28
Play from here
Copy citation
A lot of really interesting entrepreneurial activities.
00:32
Play from here
Copy citation
In fact, when I first met Matt, he was sequestered in Turkey, actually, in a
00:38
Play from here
Copy citation
hacker house where he was surrounded by a bevy of other CEOs who are all locking
00:44
Play from here
Copy citation
themselves in a room trying to solve hard AI problems together, which of
00:49
Play from here
Copy citation
course, ticked my fancy right away.
00:51
Play from here
Copy citation
So again, Matt, welcome to AI at Work.
00:54
Play from here
Copy citation
Kevin, good to have you.
00:55
Play from here
Copy citation
Good to be here.
00:56
Play from here
Copy citation
So Matt, tell us a little bit about sort of your background and your road into
01:01
Play from here
Copy citation
AI and Rapid Developers and what you guys do.
01:04
Play from here
Copy citation
Yeah.
01:04
Play from here
Copy citation
So we started the company about five years ago.
01:06
Play from here
Copy citation
We, we got our start during the no-code revolution.
01:09
Play from here
Copy citation
So no-code was this way to do visual development and that really
01:13
Play from here
Copy citation
brought the cost down to development.
01:15
Play from here
Copy citation
So instead of writing all the code, like you could just drag and drop
01:18
Play from here
Copy citation
things on a canvas and be able to build digitally that way.
01:21
Play from here
Copy citation
So we've been on this trend of like, everyone can build digitally.
01:25
Play from here
Copy citation
You should be able to build much, much faster than traditional methods.
01:28
Play from here
Copy citation
And now with AI, that's just 10 times more true.
01:32
Play from here
Copy citation
So you can build like really quickly, really fast using these tools.
01:35
Play from here
Copy citation
And so we're right at the cutting edge.
01:36
Play from here
Copy citation
We do lots of AI implementations, lots of AI consulting, architecting
01:41
Play from here
Copy citation
different types of solutions, and then we do the actual implementation
01:44
Play from here
Copy citation
for the clients that we work with.
01:46
Play from here
Copy citation
We work with folks all over the world.
01:47
Play from here
Copy citation
Our team loves some of the projects that are like really
01:49
Play from here
Copy citation
complex, really cool use cases.
01:52
Play from here
Copy citation
So it was an exciting time to be in the space.
01:53
Play from here
Copy citation
There's so many things happening.
01:54
Play from here
Copy citation
It's also kind of scary because as a leader of a development firm, like the way
01:59
Play from here
Copy citation
you do development is changing drastically.
02:01
Play from here
Copy citation
So that has a large impact on our business as well.
02:03
Play from here
Copy citation
So both exciting and scary, but good time to be alive.
02:07
Play from here
Copy citation
So you hit on no code and I find that really, really interesting because a lot
02:11
Play from here
Copy citation
of the things that we do as far as workflow engineering and workflow
02:15
Play from here
Copy citation
optimization, AI makes a lot of that much easier to do, but no code and low code
02:23
Play from here
Copy citation
approaches have been around for quite a long time.
02:26
Play from here
Copy citation
So using Zapier, Make, formerly called IntegraMAT, and then N8n sort of popped
02:32
Play from here
Copy citation
onto the scene as the first one of those that had leaned into AI.
02:36
Play from here
Copy citation
But a lot of the things that we call AI these days are actually
02:40
Play from here
Copy citation
just workflow automations.
02:43
Play from here
Copy citation
And I'd love for you to talk a little bit more about that.
02:46
Play from here
Copy citation
I don't, no code and low code is actually something we've
02:48
Play from here
Copy citation
never covered on this show.
02:50
Play from here
Copy citation
So maybe elaborate a little bit about that and your journey from low code
02:54
Play from here
Copy citation
into AI and what that looked like.
02:57
Play from here
Copy citation
I imagine for you, December 22 was kind of a, a, a really exciting time as these
03:05
Play from here
Copy citation
new tools were obviously going to make your world so much better.
03:08
Play from here
Copy citation
Yeah.
03:09
Play from here
Copy citation
So a couple of things there.
03:10
Play from here
Copy citation
First, you, we could talk about workflows versus actual like
03:13
Play from here
Copy citation
different types of AI development.
03:15
Play from here
Copy citation
So a lot of AI development is workflow.
03:16
Play from here
Copy citation
And then there's another side, which is just sort of
03:18
Play from here
Copy citation
software development in general.
03:20
Play from here
Copy citation
And that's either using AI to code, to make software or
03:24
Play from here
Copy citation
actually having like an AI product.
03:26
Play from here
Copy citation
So there's a whole range of different things we can talk about there.
03:29
Play from here
Copy citation
In terms of the progression of the no code ecosystem and the evolution of
03:33
Play from here
Copy citation
that and how that plays into AI, I would say when 2022 came out, I was
03:38
Play from here
Copy citation
like really excited, right?
03:39
Play from here
Copy citation
This technology is going to change the world.
03:40
Play from here
Copy citation
But it wasn't clear, like how much of an impact it would have on actually
03:45
Play from here
Copy citation
developing, uh, like writing code.
03:47
Play from here
Copy citation
It wasn't until like more late 2025 that it was like, wow, Lovable came out.
03:53
Play from here
Copy citation
I think it like November, 2024, it really took off in February, 2025.
03:59
Play from here
Copy citation
And for folks out there who don't know what Lovable is, it's just a platform
04:02
Play from here
Copy citation
that allows you to just chat and generate applications or websites just by chatting.
04:08
Play from here
Copy citation
No code coding experience needed.
04:11
Play from here
Copy citation
So that was a huge step change.
04:12
Play from here
Copy citation
And it wasn't until probably mid late 2025, where I started to realize like,
04:16
Play from here
Copy citation
wow, okay, the environment and landscapes changing drastically, you can
04:20
Play from here
Copy citation
really use AI to develop faster.
04:23
Play from here
Copy citation
And so what happened is before that I was kind of convinced people wouldn't
04:27
Play from here
Copy citation
want more code that it would just, this would be extra juice for the no code
04:31
Play from here
Copy citation
ecosystem where you could drag and drop things on a canvas.
04:34
Play from here
Copy citation
You could look at workflows in plain language, like Zapier and 8n, they have
04:39
Play from here
Copy citation
this like visual interface where you can see the logic steps yourself
04:43
Play from here
Copy citation
and just read it in plain English.
04:45
Play from here
Copy citation
But what I think is happening is people are less interested in those step-by-step
04:49
Play from here
Copy citation
interfaces or these canvases where you can drag and drop and really the chat
04:54
Play from here
Copy citation
interface is dominating everything.
04:56
Play from here
Copy citation
And so we'll see what happens over the next few years.
04:58
Play from here
Copy citation
Like maybe these two universes were, will merge and the AI chat based
05:03
Play from here
Copy citation
interfaces will have more canvas like functionality or the, the traditional
05:08
Play from here
Copy citation
no code ecosystem platforms will have more of a chat interface and they'll
05:12
Play from here
Copy citation
kind of converge into one kind of UX UI that everyone likes, but for now they
05:17
Play from here
Copy citation
were two different ecosystems, not a ton of overlap and I would say the AI side,
05:22
Play from here
Copy citation
the chat based interface is dominating the conversation, the airwaves and also
05:26
Play from here
Copy citation
the things we, those are the platforms we're using as well.
05:29
Play from here
Copy citation
All right.
05:30
Play from here
Copy citation
So, so I'm going to react to just a couple of pieces there.
05:32
Play from here
Copy citation
Yeah.
05:32
Play from here
Copy citation
I would say that at Ascend, one of the things that we find really frequently
05:37
Play from here
Copy citation
is we're, we're often engaged to come in and hunt around for the ways
05:41
Play from here
Copy citation
to apply AI in, in organizations.
05:44
Play from here
Copy citation
And we start turning over rocks and we start turning over rocks that the
05:48
Play from here
Copy citation
organization hasn't turned over for a long time and a lot of the time, the
05:52
Play from here
Copy citation
things that we discover aren't necessarily problems in need of AI.
05:58
Play from here
Copy citation
They're just workflow problems.
05:59
Play from here
Copy citation
And particularly as you were describing in pre-2025, the best way to go about
06:05
Play from here
Copy citation
solving a lot of those workflow problems was with these workflow builders.
06:09
Play from here
Copy citation
And, N8n was really the one that, that started to shine in that direction.
06:14
Play from here
Copy citation
But I wholeheartedly agree with you that there's been a step change, really
06:19
Play from here
Copy citation
at the beginning of 2026, end of 2025, I mean, just the pacing of all of this
06:24
Play from here
Copy citation
is completely nuts, but N8n had like a year to shine and the no code approach
06:32
Play from here
Copy citation
to using a vibe coding type tools.
06:34
Play from here
Copy citation
You mentioned Lovable, there's Repl.it.
06:37
Play from here
Copy citation
Of course, on this show, we talk a lot about Cloud Code and you suddenly
06:41
Play from here
Copy citation
didn't need those visual representations because you could basically
06:47
Play from here
Copy citation
describe what you were trying to solve.
06:49
Play from here
Copy citation
And yes, those visual representations can be really helpful if you're a process
06:53
Play from here
Copy citation
engineer and you're like breaking down every stage, but if you're a manager or
06:58
Play from here
Copy citation
an independent contributor, whatever it is, who's trying to solve a problem, the
07:03
Play from here
Copy citation
right way of looking at this in a lot of respects is what is the desired outcome?
07:08
Play from here
Copy citation
Not what are all the things that are currently being automated?
07:11
Play from here
Copy citation
Because it's almost always a mistake.
07:14
Play from here
Copy citation
If you have whatever process in your, in your organization, it's a, you know, 17
07:19
Play from here
Copy citation
different steps and Sally hands something to Joe and Joe hands something to Dave.
07:23
Play from here
Copy citation
And you're like, okay, cool.
07:25
Play from here
Copy citation
We can, we can use a workflow tool like N8N and we can automate all 17 of those steps.
07:31
Play from here
Copy citation
Well, at the end of the day, there's something that started the process.
07:35
Play from here
Copy citation
There's like a trigger, some need in the organization and there's some output.
07:39
Play from here
Copy citation
And that is what sort of delightfully most people in organizations, that's
07:44
Play from here
Copy citation
the way they think about things.
07:46
Play from here
Copy citation
They don't necessarily need to care about all 17 of those steps.
07:49
Play from here
Copy citation
And I think what I probably missed was that by being able to chat with a coding
07:57
Play from here
Copy citation
agent like that, you can basically target your desired outcome and then sort of flail
08:02
Play from here
Copy citation
through what are the necessary steps under the hood to get there.
08:06
Play from here
Copy citation
I guess, curious about your reaction to that.
08:08
Play from here
Copy citation
Yeah, a couple of things.
08:09
Play from here
Copy citation
So first, yeah, N8N is booming, you know, you get your certification.
08:13
Play from here
Copy citation
We're like, we're going to be official N8N partners.
08:17
Play from here
Copy citation
We're going to the moon.
08:17
Play from here
Copy citation
And the next thing you know, nobody cares about N8N anymore.
08:20
Play from here
Copy citation
So it's funny, like how fast this stuff is changing, just to your point.
08:23
Play from here
Copy citation
On the workflow side, yeah, you start, you dive in, you discover they have a Zapier
08:27
Play from here
Copy citation
portfolio of zaps that are like, you know, 1,000 zaps and some of them no one's
08:32
Play from here
Copy citation
touched for a long time and they're still running in the background.
08:34
Play from here
Copy citation
So there's this kind of like rat's nest that you have to dig through.
08:38
Play from here
Copy citation
Or you take the approach that you're talking about, which is you just think,
08:41
Play from here
Copy citation
what is the business problem you're trying to solve?
08:43
Play from here
Copy citation
What is it you're trying to achieve just from a business?
08:45
Play from here
Copy citation
Don't even talk about automation.
08:47
Play from here
Copy citation
Don't even talk about tech.
08:48
Play from here
Copy citation
What is it you want?
08:49
Play from here
Copy citation
Like, what are you trying to achieve?
08:50
Play from here
Copy citation
What's the output?
08:51
Play from here
Copy citation
What's the outcome?
08:52
Play from here
Copy citation
And then you can kind of do the steps you're talking about of, you know, let us
08:57
Play from here
Copy citation
worry about the steps to get there, but just describe the overall goal that
09:01
Play from here
Copy citation
you're trying to achieve.
09:03
Play from here
Copy citation
I think where some of these visual things fail too is like those steps.
09:06
Play from here
Copy citation
Sometimes, you know, if we're going, you don't need AI for everything.
09:09
Play from here
Copy citation
So like a lot of CEOs are like, I want AI.
09:12
Play from here
Copy citation
And, you know, they come up with a list of things they want and you're like,
09:14
Play from here
Copy citation
well, well, well, you know, that's not really a great use of AI for X, Y, Z
09:17
Play from here
Copy citation
reason, you don't need to automate everything.
09:19
Play from here
Copy citation
You don't need AI for every single piece of your business.
09:22
Play from here
Copy citation
Of course, there are parts that it will have a major impact for.
09:25
Play from here
Copy citation
And so where some of those visual tools fall short is sometimes you need an
09:29
Play from here
Copy citation
interface for the human to be the co-pilot, right?
09:31
Play from here
Copy citation
And so like the human can't be the co-pilot very easily.
09:35
Play from here
Copy citation
If these steps are happening through N8n or Zapier and they can't interact because
09:39
Play from here
Copy citation
there's no interface for them to interact.
09:41
Play from here
Copy citation
So sometimes like for these more complex workflows that you want to insert AI and
09:46
Play from here
Copy citation
augment the process with, you still need your human co-pilot and that's where like
09:50
Play from here
Copy citation
some of these interfaces work well.
09:52
Play from here
Copy citation
And so, yeah, that's why we're using cloud code and other types of tools to, to kind
09:56
Play from here
Copy citation
of achieve that business outcome that people want.
09:59
Play from here
Copy citation
So the related topic is the difference between a deterministic process
10:03
Play from here
Copy citation
and a probabilistic process, right?
10:05
Play from here
Copy citation
Yeah.
10:06
Play from here
Copy citation
So, so many things in business can be deterministic as in like two plus two is
10:12
Play from here
Copy citation
going to equal four every single time in Excel.
10:14
Play from here
Copy citation
And if it doesn't, something is really wrong with your Excel, right?
10:18
Play from here
Copy citation
And you really, from an efficiency perspective and from a stability and
10:23
Play from here
Copy citation
quality perspective, you want to make sure that you're being deterministic
10:27
Play from here
Copy citation
wherever you can be in the process.
10:29
Play from here
Copy citation
Using a lot of versions of these like low code, no code tools and how they've turned
10:34
Play from here
Copy citation
into AI tools and only be applying the probabilistic layer as in like where the
10:40
Play from here
Copy citation
AI needs to make sense of those outputs.
10:44
Play from here
Copy citation
That's the intelligence layer.
10:46
Play from here
Copy citation
So also from a cost management perspective, we definitely run into this, this
10:51
Play from here
Copy citation
phenomenon, sort of what you were describing with the CEO wants AI.
10:54
Play from here
Copy citation
Okay.
10:54
Play from here
Copy citation
Well, we're just going to do AI.
10:56
Play from here
Copy citation
And you end up with this spaghetti bowl of information and you're
11:00
Play from here
Copy citation
using a probabilistic layer to it.
11:03
Play from here
Copy citation
So you're using like an LLM interface on top of the spaghetti bowl of
11:07
Play from here
Copy citation
data, and it's really inefficient.
11:10
Play from here
Copy citation
And you know, one of the known weaknesses of AI is the drift and hallucinations
11:15
Play from here
Copy citation
and things like that, that happen.
11:16
Play from here
Copy citation
And you can limit a lot of that if you try and stay as deterministic as you can, as
11:22
Play from here
Copy citation
long as you can, and then apply that probabilistic layer only in the decision
11:27
Play from here
Copy citation
point and as you know, a theme lately has been costs are definitely going up.
11:32
Play from here
Copy citation
I suspect costs will continue to go up.
11:35
Play from here
Copy citation
Um, we're all going to have to be a lot more efficient about it.
11:37
Play from here
Copy citation
Yeah, definitely.
11:38
Play from here
Copy citation
I think it's best with deterministic processes, but so many things are
11:42
Play from here
Copy citation
probabilistic and as long as you have a human in the loop, like it's still, it
11:47
Play from here
Copy citation
can still speed them up if it's drafting something, drafting emails or whatever
11:51
Play from here
Copy citation
it is, or making connections through diverse, lots of data sets that a
11:54
Play from here
Copy citation
human just can't look through.
11:56
Play from here
Copy citation
There's so many things that you could still use it for, even if it's
12:00
Play from here
Copy citation
probably not deterministic.
12:03
Play from here
Copy citation
Yeah.
12:04
Play from here
Copy citation
So when, when you come into a new client, I guess first, what's sort of
12:09
Play from here
Copy citation
the, the archetype of a client for you?
12:11
Play from here
Copy citation
What, what does your, your client base look like?
12:13
Play from here
Copy citation
I think that can kind of help the audience orient a little bit.
12:16
Play from here
Copy citation
We have a lot of entrepreneurs, CEOs, you know, senior leaders, but it tends
12:22
Play from here
Copy citation
to be, I would say larger mid-market or I'm sorry, larger, small, so like big,
12:27
Play from here
Copy citation
small versus small, big.
12:30
Play from here
Copy citation
So, you know, you might have 25 employees.
12:32
Play from here
Copy citation
You might have 50 employees.
12:34
Play from here
Copy citation
Yeah.
12:34
Play from here
Copy citation
So there's sort of two main segments versus our first time founders or our
12:39
Play from here
Copy citation
startups that maybe have some initial funding and they're trying
12:42
Play from here
Copy citation
to get a product off the ground.
12:43
Play from here
Copy citation
And so they're hiring us either for the UX UI or the product consulting or to
12:49
Play from here
Copy citation
actually build the application, whether it's web, mobile or cross-platform.
12:53
Play from here
Copy citation
So that's one segment.
12:54
Play from here
Copy citation
The main segment though, is working with CEOs, let's say businesses from
12:58
Play from here
Copy citation
like 20 million up to $500 million of revenue, understand what AI
13:02
Play from here
Copy citation
could be doing for their business.
13:03
Play from here
Copy citation
So a couple of different aspects are one is education, whether it's getting
13:07
Play from here
Copy citation
their workforce change, doing a workshop, hackathon to get people kind of
13:11
Play from here
Copy citation
understanding what these tools are and how to use them, or actually architecting
13:15
Play from here
Copy citation
a system, doing some consulting.
13:18
Play from here
Copy citation
We always start the consulting from a business side first before
13:20
Play from here
Copy citation
we dive into the tech side.
13:22
Play from here
Copy citation
But if there is some sort of tech component, then we can proceed with
13:25
Play from here
Copy citation
like the implementation of if they don't have an engineering team themselves.
13:28
Play from here
Copy citation
So typically do you go in, which is it?
13:31
Play from here
Copy citation
Is it a push or a pull?
13:32
Play from here
Copy citation
Like the CEO has a laundry list of things or the CEO says, help me
13:36
Play from here
Copy citation
figure out my laundry list of things.
13:38
Play from here
Copy citation
It depends.
13:38
Play from here
Copy citation
Like in the old economy, it's more like, Hey, here's my like general ideas.
13:42
Play from here
Copy citation
Can you help me understand and build a framework for me to think how to think
13:46
Play from here
Copy citation
about AI and where it can be implemented in my business, other times for more
13:50
Play from here
Copy citation
tech savvy businesses, or maybe they have some engineering talent.
13:54
Play from here
Copy citation
They're like, Hey, we've been mulling over this idea.
13:57
Play from here
Copy citation
We have a prototype, but we kind of need a professional come in.
13:59
Play from here
Copy citation
And, you know, now that we have a proof of concept to actually make
14:02
Play from here
Copy citation
this very robust and scalable versus our team's too busy, or it just doesn't
14:06
Play from here
Copy citation
have the skills to be able to do that.
14:08
Play from here
Copy citation
Okay.
14:08
Play from here
Copy citation
Yeah.
14:09
Play from here
Copy citation
So a little bit of both.
14:10
Play from here
Copy citation
So what, what sort of solutions are you building for folks?
14:13
Play from here
Copy citation
So if you, if you had to compartmentalize it, like, you know, what's sort of the
14:17
Play from here
Copy citation
go-to things that you are coming up with, or your clients are coming
14:22
Play from here
Copy citation
up with for first areas of focus?
14:25
Play from here
Copy citation
I would say there's a huge range.
14:26
Play from here
Copy citation
I haven't seen, spotted too many distinct trends yet.
14:30
Play from here
Copy citation
I think one of the most distinct trends is like, everybody wants a brain.
14:33
Play from here
Copy citation
They're like, Hey, can like this thing, can you just build me a system that knows
14:36
Play from here
Copy citation
every single thing about my business and I can just talk to it and query it and ask.
14:39
Play from here
Copy citation
questions. And then can you add some agents so I can just start doing stuff for me? So
14:44
Play from here
Copy citation
obviously that's very broad. But I think like the brain, the concept of having like everything
14:49
Play from here
Copy citation
in one system that can query everything that's happening, whether it's all your Slack or
14:55
Play from here
Copy citation
all the emails at the company or all the videos. We actually have this internally for our own
15:00
Play from here
Copy citation
company and it runs like a ton of stuff. So we have about like 250 clients at any given
15:05
Play from here
Copy citation
moment. And we have every single email, every single Slack message, every single call transcript
15:12
Play from here
Copy citation
is loaded into this AI brain system and it actually segments it based on project. And
15:18
Play from here
Copy citation
so we can see what's the status of the project, if there's any escalations, if anything's
15:22
Play from here
Copy citation
off track, we can see moments where the client is happy and we could ask for a review or
15:28
Play from here
Copy citation
they're having a problem unrelated to our business and they just have another problem
15:33
Play from here
Copy citation
they need help solving and we can connect them to a vendor that can help them.
15:36
Play from here
Copy citation
So this type of system allows us to have a lot of vigilance and monitoring and flagging
15:42
Play from here
Copy citation
important different types of alerts, like unlike we've ever been able to do. So that's
15:46
Play from here
Copy citation
why I'm excited about our types of businesses, because we can scale this in a way that's
15:51
Play from here
Copy citation
never, you know, traditionally professional service firms have struggled to scale because
15:55
Play from here
Copy citation
of the human factor. But that's an example of something we have, something that we see
15:59
Play from here
Copy citation
quite commonly asked for across different types of businesses, regardless of sector
16:03
Play from here
Copy citation
or industry.
16:04
Play from here
Copy citation
So we see the same thing. We're doing a lot of projects in that respect. And I do find
16:10
Play from here
Copy citation
it pretty interesting as far as a market hole in that you have like Glean. Have you heard
16:16
Play from here
Copy citation
of Glean?
16:17
Play from here
Copy citation
Yeah.
16:18
Play from here
Copy citation
So, so the, this is an enterprise solution that does this and it does a really pretty
16:23
Play from here
Copy citation
good job of it, to be honest. Like I want to like poopoo like enterprise solutions.
16:28
Play from here
Copy citation
It's not cheap for sure, but they're doing a pretty good job of pulling in a lot of different
16:32
Play from here
Copy citation
pieces and they're allowing people within these larger organizations to operate with
16:38
Play from here
Copy citation
a lot more flexibility because the system, the brain also has guardrails that are built
16:45
Play from here
Copy citation
around it. I suspect given my experience with sort of mid-market type companies, the brains
16:51
Play from here
Copy citation
don't have as much guardrailing in place, which is okay with the clients. They're understanding
16:56
Play from here
Copy citation
some of the trade-offs as far as information sharing and some of those risks it entails,
17:01
Play from here
Copy citation
but that's a complete non-starter for others.
17:03
Play from here
Copy citation
And I'm curious on your take as far as why isn't there a good solution for this that
17:10
Play from here
Copy citation
has popped up yet? Because it is glaringly obvious. It's also very hard to do. And, you
17:16
Play from here
Copy citation
know, both of us are monetizing it. So, you know, okay, by all means market don't do this.
17:22
Play from here
Copy citation
But what do you think are the, why hasn't the market come up with a solution that cleanly
17:30
Play from here
Copy citation
builds a more mere mortal version of Glean?
17:33
Play from here
Copy citation
Yeah. I think what we discovered when building it ourselves is like the fidelity of these
17:38
Play from here
Copy citation
systems drops drastically once you get too much information into it. So if you have 5,000
17:44
Play from here
Copy citation
employees, good luck. Like this is going to be, that's your way too complex and way too
17:48
Play from here
Copy citation
much information for, to be kind of organized by these systems. At this point, everything
17:52
Play from here
Copy citation
I think could change in the future. But for us, the secret was when we can figure out
17:57
Play from here
Copy citation
a way to segment the data, like for example, projects, if we have 250 clients, we create
18:03
Play from here
Copy citation
250 little mini brains and those 250 mini brains just monitor everything for that one
18:09
Play from here
Copy citation
client, the slacks, the emails, the call transcripts just for that one client.
18:12
Play from here
Copy citation
And then the fidelity is extremely high because it's able to, its brain is big enough to,
18:18
Play from here
Copy citation
to see thousands of slack messages, emails, and call transcripts for that one client.
18:23
Play from here
Copy citation
But it's not capable of seeing billions of lines of text across all our clients. And
18:28
Play from here
Copy citation
so when we think about how to make this work, we try to figure out a way to cut it so that
18:33
Play from here
Copy citation
the brains are sort of, you know, in smaller batches. And that makes the fidelity much,
18:38
Play from here
Copy citation
higher for the result, for the alerts, for the querying, you get much, much better feed
18:44
Play from here
Copy citation
messages back from the system.
18:47
Play from here
Copy citation
So I'll say on sort of a continuum, like the, your, what you're talking about, I would say
18:51
Play from here
Copy citation
sort of is pegged to the right side of the continuum of it being pretty robust on a per
18:56
Play from here
Copy citation
client basis. Obviously it does have caps as far as size. These things can also be really
19:01
Play from here
Copy citation
expensive to run.
19:02
Play from here
Copy citation
Yes, yes.
19:03
Play from here
Copy citation
That it's something that you need to be very careful with because you're rapidly falling
19:08
Play from here
Copy citation
out of the, the, the just running a Claude Max plan or running a, you know, a hundred
19:13
Play from here
Copy citation
dollar a month GPT plan or whatever it is. And all of that's working. As soon as you
19:19
Play from here
Copy citation
start building these things, and I can't emphasize this enough, you fall out of those
19:23
Play from here
Copy citation
monthly plans and now you're in API land and the API, the cost per call continues to go
19:31
Play from here
Copy citation
up as these companies are heading for IPO. So you always want to be cautious about that.
19:36
Play from here
Copy citation
And it's still worth it. I mean, when we talk expensive, if these things are well built,
19:41
Play from here
Copy citation
it's, it's not that bad given the amount of functionality that you get out of them.
19:47
Play from here
Copy citation
But then on the other side of the continuum, we've talked a lot on this show, as far as
19:52
Play from here
Copy citation
building your own personal chiefs of staff and being able to use things like Claude Cowork
19:58
Play from here
Copy citation
or GPT agents in order to, to get some of those pieces out of the material. So I don't
20:05
Play from here
Copy citation
want anybody listening to like freak out that, Oh my gosh, I'm going to have to spend tens of
20:09
Play from here
Copy citation
thousands of dollars. It's going to be all of this development work. And then we're going to
20:12
Play from here
Copy citation
have this high operating costs to run these things. Like there's a starting point, which is
20:19
Play from here
Copy citation
essentially simply tie tying in the agent structures that are already part of your
20:24
Play from here
Copy citation
Claude Max plan. And I'm just, I tend to be a Claude guy. So I tend to go in that direction,
20:30
Play from here
Copy citation
but you can get a long way along this continuum by simply tying in Claude or GPT via connector
20:38
Play from here
Copy citation
to say your project management tool or to your email or to your calendar and be able to get
20:43
Play from here
Copy citation
a lot of insights from that. So if you're a one, two, three, four type person company,
20:49
Play from here
Copy citation
like you don't have to get crazy about this. I guess sort of react to that.
20:53
Play from here
Copy citation
Yeah, no, I think the other thing that we see commonly people coming to us, CEOs are like,
20:58
Play from here
Copy citation
Hey, I got this thing, but it costs me $10,000 a month or something. And we go in and we see,
21:03
Play from here
Copy citation
it's just the folks that we're building that aren't as proficient and haven't had enough reps
21:08
Play from here
Copy citation
yet. And in general, there's just not that many people in the market that have experience actually
21:12
Play from here
Copy citation
building real things. And so some of the work that we're doing is just optimizing the stuff.
21:16
Play from here
Copy citation
Like, Hey, well, you're using like a blowtorch to light a cigarette or something, right? Like,
21:22
Play from here
Copy citation
you don't need that. Like let's stage it. Let's do like the initial filter with the
21:26
Play from here
Copy citation
cheapest models possible. You know, you don't even have to use the, the premium models,
21:30
Play from here
Copy citation
like the other, some of the open source models could be good. Like there, there's a lot of
21:35
Play from here
Copy citation
models out there that are much cheaper for the simpler tasks that, you know, you want to do some
21:39
Play from here
Copy citation
initial layer of filtering for your brain. And then you, then you query a more complex with like
21:45
Play from here
Copy citation
the higher end models. So there's ways to kind of cut the cost down to be more reasonable. And
21:49
Play from here
Copy citation
that's actually like a fair amount of work we see coming in. Yeah. I think in general, I think the
21:55
Play from here
Copy citation
cost of the models will come down. So I don't know, you know, I know they're all getting ready
21:58
Play from here
Copy citation
for IPO. We'll see, but there is the other thing is like hosting things locally. You know, like
22:04
Play from here
Copy citation
there's some companies that are really worried about privacy and sending all their stuff to cloud
22:08
Play from here
Copy citation
their trade secrets. I think that's more for much, much bigger companies, but like people are still
22:12
Play from here
Copy citation
thinking about that and maybe they're in a HIPAA compliant healthcare industry and they need to be
22:16
Play from here
Copy citation
aware of like what model they're using or they have to use it locally. So there's all those types of
22:21
Play from here
Copy citation
governance security and cost issues that we see all the time too. Yeah. I, I, it's funny because
22:27
Play from here
Copy citation
if you have a max plan, like, I don't know about you, but I tend to just use Opus all the time.
22:32
Play from here
Copy citation
Yeah. Same. Because why would I want the dumber model? And it does make you pretty inefficient.
22:38
Play from here
Copy citation
If you, if you take that same logic and you apply it into, into an app build, and there is a lot of
22:45
Play from here
Copy citation
nuance in deciding which model to use, where, and for what to get the highest level of efficacy for
22:51
Play from here
Copy citation
the lowest cost. And that's a little bit of an art for, so for those of you who are listening
22:56
Play from here
Copy citation
and you have people who are working on this, or you yourself are working with it, you always want
23:01
Play from here
Copy citation
to play that game. And you also want redundancy that's built into the system as well. Like as
23:06
Play from here
Copy citation
much as I like Anthropic, it goes down all the time. And if you're totally dependent on a single
23:11
Play from here
Copy citation
model, this is going to be a problem going forward that some, the CIO or the CTO is going to wake up
23:18
Play from here
Copy citation
one day and realize that like their entire stack has basically collapsed just because of one point
23:26
Play from here
Copy citation
of failure because Anthropic couldn't handle whatever the load is. So we, we call this, this,
23:32
Play from here
Copy citation
this like rejiggering, we actually call it AI plumbing. And I think it's actually going to be
23:36
Play from here
Copy citation
like a thing where, why do you call the plumber? You call the plumber because the pipes broke,
23:41
Play from here
Copy citation
right? And, you know, nothing, nothing against people who were building in January one, you know,
23:47
Play from here
Copy citation
2023. But as you said, nobody really knew what they were doing at that time. And a lot of things
23:54
Play from here
Copy citation
have been built over the last three years that you'd never build it that way in the first time.
24:00
Play from here
Copy citation
And now it's fragile and it needs to be reconstructed. And I don't actually see spinning
24:05
Play from here
Copy citation
this up as a product line, but like, I could totally see people like us running businesses
24:11
Play from here
Copy citation
that are the AI plumbers that clean up, go back and they rip off the duct tape. They, you know,
24:17
Play from here
Copy citation
they like reconnect things in a, in a, in a, in a more efficient sort of way going forward.
24:24
Play from here
Copy citation
Yeah. Like I had a post, it was like vibe coding cleanup specialists. And that's like a new role,
24:30
Play from here
Copy citation
you know, like there's tons of stuff like that. It also begs the question, Kevin, I'd be interested
24:34
Play from here
Copy citation
to know what you think of this. Like I was doing a, I was doing a speech somewhere for like a
24:40
Play from here
Copy citation
conference I was at and somebody asked like, Hey, I'm scared to invest in AI because I feel like
24:45
Play from here
Copy citation
everything's changing so fast. You know, let's say I do this for my accounting system, but you
24:51
Play from here
Copy citation
know, someone's just going to come up with a product that does exactly what I just spent a
24:54
Play from here
Copy citation
hundred K to build or whatever it was. Right. So how do you think about like, Hey, where should I
25:00
Play from here
Copy citation
invest or where should I wait? Because I, sometimes it's not always great to be first.
25:05
Play from here
Copy citation
Sometimes, especially for small businesses that don't have the free money tree, you know, every
25:10
Play from here
Copy citation
dollar counts, you know, should they be, how should, what framework should they have about
25:15
Play from here
Copy citation
where to invest and where to wait to see if someone productizes something off the shelf for
25:19
Play from here
Copy citation
them? I think it's a, it's a great question. And the way that I usually put it to people is you
25:25
Play from here
Copy citation
want to build something. If it creates a competitive advantage for you, that there's
25:32
Play from here
Copy citation
nothing on the market that really solves the problem in the way that you need to solve the
25:36
Play from here
Copy citation
problem. And it's going to make a difference to the way that you go to market. The next category
25:42
Play from here
Copy citation
are, are, are the by category and the by category are the things that scratch an itch that are
25:48
Play from here
Copy citation
generally adjacent to the rest of your business. That's things that, that are important, but
25:55
Play from here
Copy citation
they're not necessarily urgent and they don't necessarily have to be perfect for the gears of
26:00
Play from here
Copy citation
the business to keep turning. And then the last category is basically wait, because if it's not,
26:06
Play from here
Copy citation
if, if it's not in that, I need to build it category and you can't really find it in the
26:12
Play from here
Copy citation
by category. There are probably two Estonians with a dream in their garage who are building
26:18
Play from here
Copy citation
it right now in whatever vertical you're in, and it's going to pop on the market. However, however,
26:22
Play from here
Copy citation
however, I will say that, that I am kind of pivoting a little bit in the way that I'm thinking
26:28
Play from here
Copy citation
about this because just by the nature of SaaS, I mean, I don't, what do you, do you Salesforce,
26:34
Play from here
Copy citation
HubSpot, what's sort of your CRM of choice? Oh, we're actually replacing it with like a AI native
26:39
Play from here
Copy citation
CRM. Okay. So, so there you go. But anytime you open a CRM, you can see the effect of broad
26:46
Play from here
Copy citation
market SaaS, because it has to be everything to everyone. And then it ends up being kind of
26:50
Play from here
Copy citation
nothing to anyone. And so you, you're a sales business, but that's not actually your business.
26:55
Play from here
Copy citation
So being able to build a customized SaaS internally or CRM internally is something
27:02
Play from here
Copy citation
that would have fallen into that, that, that by category before, but I promise you, you're never
27:08
Play from here
Copy citation
going to find a CRM that's going to scratch all the issues that you want. And because you have
27:12
Play from here
Copy citation
the capacity to build it, it can make sense, particularly when there's a big ticket that's
27:18
Play from here
Copy citation
associated with it. I mean, you've got a lot going on. So your previous CRM bill was material. It was,
27:24
Play from here
Copy citation
you know, probably five to $10,000 a month, I would guess, given your size. And that adds up
27:30
Play from here
Copy citation
internally, you're not going to spend 120 grand developing a solution that works straight for you.
27:37
Play from here
Copy citation
It becomes a matter of prioritization, I would say. And this I run into a lot. In fact, this is
27:42
Play from here
Copy citation
happening right now with a client who I dearly love, but they are doing amazing things and
27:47
Play from here
Copy citation
changing their product and making huge product investments that are going to totally differentiate
27:52
Play from here
Copy citation
them. But that has their team so stretched thin that when they're confronted with, say something
27:59
Play from here
Copy citation
that's like a $400,000 like project, they just don't have the time for it. And they don't have
28:05
Play from here
Copy citation
the training or the tools or the staff that can like actually go in and solve that problem. But
28:10
Play from here
Copy citation
they probably should be building the $400,000 solution as well. And really, that's where people
28:16
Play from here
Copy citation
like us come in. That's where, you know, they should be focused on the rocket science per their
28:20
Play from here
Copy citation
business. That's where their hiring should be. That's where they're going to find the best and
28:24
Play from here
Copy citation
brightest who are going to bring AI to the product. But as far as executing in the rest of the
28:30
Play from here
Copy citation
organization, for at least a time here, you really do need organizations like ours to be able to solve
28:36
Play from here
Copy citation
those problems. Yeah, I think let's use ERPs as another example or another category that we see.
28:42
Play from here
Copy citation
We talked about AI brains being like such an obvious like ask in the market from lots of CEOs.
28:48
Play from here
Copy citation
I think ERPs are another example because they're just so old and antiquated and terrible to work
28:53
Play from here
Copy citation
with and expensive and the service cost on top of the software cost is enormous. So I think that's
28:59
Play from here
Copy citation
like a good example to your point of don't wait for that. ERPs are always so horizontal, so
29:06
Play from here
Copy citation
generalized, and there's so many things and agents that you can build off an ERP that are accustomed
29:12
Play from here
Copy citation
to your business. That's an example. Maybe it's not the CRM, but maybe an ERP is an example of
29:17
Play from here
Copy citation
something that you really want to own yourself and
29:19
Play from here
Copy citation
invest in. Examples that are bad, I think, are like, you know, 2023, everyone making
29:24
Play from here
Copy citation
custom voice agents versus just waiting for VAPI, where now it's like super easy plug
29:29
Play from here
Copy citation
and play. Same thing with chatbots. Everyone's making their custom chatbots in 2023. And
29:34
Play from here
Copy citation
now there's like a million just off the shelf tools to do your chatbot and your knowledge
29:37
Play from here
Copy citation
base off. I think accounting, anything with some types of invoicing will eventually fall
29:43
Play from here
Copy citation
into that. I think there's a lot of people working on bookkeeping and accounting tools
29:46
Play from here
Copy citation
because it's so labor intensive and it's so easy to kind of augment that what they
29:50
Play from here
Copy citation
are. But yeah, I think those, just to add some examples to the framework you presented
29:54
Play from here
Copy citation
our listeners.
29:55
Play from here
Copy citation
So I find that I have sort of a visceral reaction now to SaaS platforms that are desperately
30:02
Play from here
Copy citation
trying to say about the stay alive by like walling off their data or figuring out, you
30:08
Play from here
Copy citation
know, like, I mean, they've got to do what they've got to do. But now that we have the
30:13
Play from here
Copy citation
flexibility to basically build around it, that's another category where if the platform
30:19
Play from here
Copy citation
isn't going to give me the access that I need to do the things that I need to fully automate
30:26
Play from here
Copy citation
the rest of my process, that moves it up in my ranking as far as something that I'm likely
30:31
Play from here
Copy citation
to build versus buy. And I am looking at you like accounting platforms out there. Like
30:37
Play from here
Copy citation
it just makes me die inside to interact with some of these things.
30:41
Play from here
Copy citation
Now there are pieces of that puzzle that I'm never going to touch. There's no way you're
30:45
Play from here
Copy citation
going to have, I'm going to try and recreate like Stripe for a transaction.
30:48
Play from here
Copy citation
I was just about to say, we're not, don't be a payment processor, don't be able to strike.
30:53
Play from here
Copy citation
Don't be a like gusto. I hate paying like payroll. It's expensive, right? To, to, to
30:59
Play from here
Copy citation
do payroll and it looks basic, but they're dealing with, you know, 50 states of regulation
31:05
Play from here
Copy citation
and 200 countries.
31:06
Play from here
Copy citation
Like yeah, anything compliance heavy, stay away.
31:09
Play from here
Copy citation
And that actually lends itself a little bit to an investment lens, right? Like, like,
31:13
Play from here
Copy citation
you know, you can be doing something that's, that, that's really seems like it's obvious
31:17
Play from here
Copy citation
to be recreated like payroll or state tax compliance. So there's like Avalara that does
31:23
Play from here
Copy citation
state tax compliance out there. And there are other platforms that do that, but they're
31:29
Play from here
Copy citation
taking on a lot of that liability in a way that if you do it wrong, you could actually
31:34
Play from here
Copy citation
go to jail. So, you know, those that, that gives them a barrier.
31:39
Play from here
Copy citation
Now, what that experience looks like, like if I never have to go into gusto, the play
31:44
Play from here
Copy citation
payroll plot processor ever again, that day will be too soon, but I need them to do the
31:50
Play from here
Copy citation
thing that they do best. That may not sit there, suit there in their stakeholder needs
31:57
Play from here
Copy citation
or their growth needs or their market needs or whatever it is, because at the end of the
32:00
Play from here
Copy citation
day, they're a regulatorily compliant structure that is headless, that I do all the cool stuff
32:06
Play from here
Copy citation
that I want to, that, that isn't a very sexy business. And it certainly doesn't require
32:10
Play from here
Copy citation
10,000 employees to run that business. So, you know, forward looking on, on, on what
32:15
Play from here
Copy citation
that turns into, I think is really interesting. And if you are one of those companies, if
32:20
Play from here
Copy citation
you are a SaaS company, you need to be paying a lot of attention to what I just said, like
32:26
Play from here
Copy citation
how are, how are people like Matt and I and our clients, how are we looking at your platform
32:33
Play from here
Copy citation
as far as the pieces we need? And the more you try and cram me into your UX, the more
32:39
Play from here
Copy citation
likely I am to try and just do it myself. Yeah. I just called my financial advisor,
32:45
Play from here
Copy citation
maybe I think November, 2025 or December, 2025 was a huge step change for Agentic AI.
32:52
Play from here
Copy citation
And since then I called my financial advisor. I'm like, I don't know what's going to happen
32:55
Play from here
Copy citation
to SaaS. I've always told you to go heavy on SaaS. I'm telling you now, let's maybe diversify,
33:01
Play from here
Copy citation
pull out a little bit. As somebody who's in software every single day, I'm not saying I'm
33:06
Play from here
Copy citation
right, but I'm saying for sure, SaaS, SaaS valuations are going down. Their moat is
33:10
Play from here
Copy citation
disappearing, which their moat was like the cost of development, which is heading towards zero
33:14
Play from here
Copy citation
and the migration costs, which all these agents, it's no one's talking about this, but in a few
33:18
Play from here
Copy citation
years, you're going to be able to click a button and say, copy all my Salesforce configuration
33:22
Play from here
Copy citation
settings, all my database, everything from Salesforce over to this new CRM and migration
33:27
Play from here
Copy citation
costs is going to head to zero. And I think in the next few years, but no one's talking about it.
33:31
Play from here
Copy citation
So anyways, I'm pulling out, I think the only SaaS product, I even like the LLMs. Kevin,
33:36
Play from here
Copy citation
I'm not really sure. Like I know everyone's talking about bubble. I don't think it's a
33:39
Play from here
Copy citation
bubble. I'm not like doomed to stay or about that, but I just think they're in a commoditized
33:44
Play from here
Copy citation
industry. And I'm not sure if they're really that good as a businesses, like just as a business,
33:50
Play from here
Copy citation
like, it seems like lots of people can build LLMs and lots of people can have data centers
33:55
Play from here
Copy citation
and I'm not sure it's a winner take all market or monopoly. And that's usually the type of
33:59
Play from here
Copy citation
companies you want to invest in monopolies. Right. So yeah, I'm, I'm even hesitant about
34:05
Play from here
Copy citation
the LLMs out there. So yeah, I'm not sure what the future holds, but that's my quick,
34:09
Play from here
Copy citation
quick take on that. I got an idea for you though. Let me bounce it around with you.
34:13
Play from here
Copy citation
If you've got a minute. Okay. So armless, like the ZRP idea. One thing we're floating around
34:20
Play from here
Copy citation
doing a campaign around is, Hey, let us build some of those custom software for you. And we
34:26
Play from here
Copy citation
actually won't charge for the build. We'll charge for the payment side. So we'll take, instead of
34:32
Play from here
Copy citation
you, you're already paying 3.2% because you're using an expensive, let's say ServiceDyne.
34:37
Play from here
Copy citation
You're already using ServiceDyne. It's super expensive to have ServiceDyne, super expensive
34:42
Play from here
Copy citation
to get the implementation help and the service help. And then on top of that, they take like a
34:46
Play from here
Copy citation
pretty big chunk of your, all your revenue. Why don't you just let us build your ServiceDyne for
34:51
Play from here
Copy citation
you? We'll be the service provider and we'll maintain the system and it's yours. You own it.
34:55
Play from here
Copy citation
And we'll just make our money off the 3.2% that ServiceDyne already charges you. So like, I wonder
35:01
Play from here
Copy citation
if that could become a model in the future. What do you think, Kevin? Well, I think realignment
35:05
Play from here
Copy citation
of pricing is something that's coming our way. So we do, we do some engagements that we're,
35:10
Play from here
Copy citation
we're basically helping SaaS companies figure out or service companies like accountancies and things
35:16
Play from here
Copy citation
figure out like what is their new pricing model? Like, you know, people like us are like output,
35:20
Play from here
Copy citation
output, output, output. Right. But that's kind of hard to figure out. Like, I know, I think the
35:26
Play from here
Copy citation
world's good. I think that's good for clients. It's good for us to be accountable, but it's like,
35:31
Play from here
Copy citation
I want to do that, but it's so hard to measure and actually do it. Right. Absolutely. And,
35:35
Play from here
Copy citation
and there is just this like optics of it's a $200 a month, you know, subscription into the SaaS and
35:42
Play from here
Copy citation
it's doing things and you never really do that arbitrage math of how much am I on gross. You
35:49
Play from here
Copy citation
know, you're getting stuff out of it, but if it starts charging you on an incremental basis,
35:54
Play from here
Copy citation
now you have this like variable cost factor that you're trying to figure out and trying to manage
35:59
Play from here
Copy citation
against. And it's going to just drive controllers completely crazy as they're trying to figure this
36:04
Play from here
Copy citation
stuff out and it's going to be all over the board. So, you know, it's something I I've said
36:09
Play from here
Copy citation
a number of times here is it's like, you know, look at all of your vendors. If you want to,
36:13
Play from here
Copy citation
if you save some money in your organization, who is charging you by the hour as opposed to
36:19
Play from here
Copy citation
by the project. And at the very least you should be applying some pressure to them such that they
36:25
Play from here
Copy citation
need to start thinking about this in terms of deliverables. Like think about lawyers, like what
36:30
Play from here
Copy citation
is a, what is a trust worth to do? Like, I think that's pretty easy. It's much harder if it's like
36:36
Play from here
Copy citation
a trial and it's built on billing, but even internally we have this question. I was just
36:41
Play from here
Copy citation
going to say, Kevin, that sounds all wonderful, but like, how do we even do it internally when
36:46
Play from here
Copy citation
your team is also paid in a way that it's like, how do you, how do you make sure the output's fair?
36:50
Play from here
Copy citation
That's the hard part is like, is this ticket in the development world, we go off tickets,
36:54
Play from here
Copy citation
milestones and story points. And we try to like, you know, indicate what the size of the workload
37:00
Play from here
Copy citation
is for that's that milestone or that's that ticket for via story points, but it's still a subjective
37:05
Play from here
Copy citation
process, you know, and now it's affecting someone's pay. And yeah, our best, our best,
37:11
Play from here
Copy citation
like workflow engineer and automation people, they're like, wait a second. If I'm getting
37:15
Play from here
Copy citation
paid by the hour, you're actually disincentivizing me to create scaled products that, that don't
37:23
Play from here
Copy citation
require my input because although we have plenty of business, like it's still finite if you, you,
37:28
Play from here
Copy citation
you keep scaling it. And you know, these people's jobs are absolutely not in any danger. Right.
37:33
Play from here
Copy citation
But they're like, you're disincentivizing me that I'm supposed to look, work longer on this.
37:39
Play from here
Copy citation
So we're, we're, we're, we're trying to figure that out. And it, I think it, it actually requires
37:44
Play from here
Copy citation
a decent amount of transparency in the organization such that everybody understands what the expected,
37:51
Play from here
Copy citation
like, you know, what's the IRR, the internal rate of return. You want to have a solid gross margin.
37:58
Play from here
Copy citation
And then you want to encourage the team to out that gross margin such that they can share in it.
38:05
Play from here
Copy citation
And, you know, as a greedy, greedy capitalist that puts a cap on your, your individual,
38:11
Play from here
Copy citation
like engagement contribution, but aligns incentives with the team such that even if
38:17
Play from here
Copy citation
there are non-equity participants, they get to participate in that success of the company.
38:23
Play from here
Copy citation
And the company is still getting its nut as far as the number.
38:27
Play from here
Copy citation
Yeah. Yeah. I watched you rejoined her on that. I do have some thoughts on your,
38:31
Play from here
Copy citation
your, your, your transaction model too.
38:35
Play from here
Copy citation
Yeah. Yeah. I think, I mean, I think the world will be better if everything's aligned for
38:39
Play from here
Copy citation
output versus, you know, I mean, didn't SpaceX do this, right? Like, you know, they disrupted
38:44
Play from here
Copy citation
the entire defense industry because they said, Hey, we're going to, you're going to pay us based
38:49
Play from here
Copy citation
on the outcome, not on the, you know, by time and materials, which historically for the last
38:54
Play from here
Copy citation
hundred years, that's how the defense complex has been paid. Right. So, and look at them,
38:59
Play from here
Copy citation
they're now a $2 trillion company. And so when you align incentives like that,
39:03
Play from here
Copy citation
it creates a lot of value for others and it allows you to capture value too.
39:06
Play from here
Copy citation
But it's not very predictable from a financial perspective. So we're both MBAs, right? So.
39:13
Play from here
Copy citation
Don't tell people that. Come on.
39:14
Play from here
Copy citation
They're not going to listen to us.
39:15
Play from here
Copy citation
Yes. No, we're coding geeks. That's all we do all day long.
39:20
Play from here
Copy citation
All right. Dirty secrets out, everybody. Kevin and I have an MBA. Don't think about some of us.
39:28
Play from here
Copy citation
So, but it becomes really hard from a predictability perspective. And when it's nice
39:34
Play from here
Copy citation
for us to say, okay, the shawl should be output based, but when you go to a client and the client
39:39
Play from here
Copy citation
says, well, how do I set my budget? Well, I, it's hard for me to tell you how to set your budget
39:44
Play from here
Copy citation
because it's all output based and surely that's valuable to you. But the whole organizational
39:49
Play from here
Copy citation
structure on their side has been built around financial controls and budget predictability
39:54
Play from here
Copy citation
and whatever. And then even though whatever you're doing is ideally creating a lot of economic value
40:01
Play from here
Copy citation
in larger organizations, as soon as you blow through that budget, like all kinds of flags
40:06
Play from here
Copy citation
get thrown and approval processes. Yeah. So they want predictability and being able to do
40:13
Play from here
Copy citation
both output and have predictable costing is, is something that I don't think that we've figured
40:19
Play from here
Copy citation
out yet. Great point. I'm actually going to write that point down because that is brilliant.
40:25
Play from here
Copy citation
Yeah, but it, it, it, it matters as you're, you're, you're talking through repricing about
40:29
Play from here
Copy citation
all of this because it's really easy for guys that are like us to say, Hey, let's just like
40:33
Play from here
Copy citation
disrupt the entire organization and let's just break it. And you know what? I actually do have a
40:38
Play from here
Copy citation
couple of clients who are game for that. And it's, it, it really is a gut check to be like, we're
40:44
Play from here
Copy citation
just going to redo everything in the organization. And it takes a special brand of leadership.
40:48
Play from here
Copy citation
It takes people management and change management in addition to the tech to make all of the things
40:53
Play from here
Copy citation
work. But most companies don't want to do that. They just want, they want the things they're
40:58
Play from here
Copy citation
doing to continue to work. It's not really in their cards to try and replace their accounts
41:05
Play from here
Copy citation
receivable staff. Like why, why would we do that as opposed to just making it easier, but.
41:13
Play from here
Copy citation
You know, we're yet positing that you should do that because that disruption creates the
41:18
Play from here
Copy citation
opportunity for rebuilding the organization in a way that fundamentally makes more sense.
41:22
Play from here
Copy citation
I think a lot, well, surprisingly, I think, I don't know if it's surprisingly or not,
41:27
Play from here
Copy citation
but I was talking to one of the large AI platforms out there. I won't name it, but
41:33
Play from here
Copy citation
we have a meeting with them every two weeks and they were talking about how much,
41:38
Play from here
Copy citation
how fast adoption has been in the mid-market versus the larger, the company just way more
41:44
Play from here
Copy citation
complex, way slower, not embracing it, very scared, skeptical. And so like they're actually
41:50
Play from here
Copy citation
having a ton of success with small businesses and mid-market businesses because they're just
41:54
Play from here
Copy citation
so much more nimble. They can adopt these. And some of the leaders are still the founders who
41:59
Play from here
Copy citation
have that are cut from that strain of like, yeah, just burn it all down. Let's try something new,
42:03
Play from here
Copy citation
or just like a little more risk taking personalities versus your corporate CEOs who,
42:08
Play from here
Copy citation
you know, were brought up the corporate ladder to lead the company versus the founder entrepreneur
42:12
Play from here
Copy citation
who started it and is leading it, you know? So I, you know, I'd love your take on the,
42:16
Play from here
Copy citation
the other issue that's been floating around, which is, is, is job displacement. And again,
42:21
Play from here
Copy citation
we try and stay away from politics here, but I happen to have been listening to an all-in
42:26
Play from here
Copy citation
podcast this morning and they're all like really kind of relaxed about it. They,
42:31
Play from here
Copy citation
they just blaming all cut cuts on, on AI washing, you know, overhired. So of course this is just AI
42:39
Play from here
Copy citation
washing and they're just so, so yeah. And, and yet I look at my client base and I mean,
42:44
Play from here
Copy citation
whether they're saying it overtly or not, that's what they're after.
42:48
Play from here
Copy citation
I think, yeah. So a couple of things on this first meta level, like I'm definitely worried
42:54
Play from here
Copy citation
about like the, the labor market. You know, I love history. I've read lots of
42:58
Play from here
Copy citation
history. Whenever I get free time, it's like a fun hobby and topic for me.
43:02
Play from here
Copy citation
And generally I would say, you know, there's nothing new under the sun. Our society has
43:08
Play from here
Copy citation
seen disruptive changes before people have trouble rescale skilling, but it does happen
43:13
Play from here
Copy citation
and new jobs to be done will be created that we can't even imagine now. Okay. I gotcha.
43:18
Play from here
Copy citation
That's happened many times throughout history. I think the problem with this change is it's
43:22
Play from here
Copy citation
just so fast and money moves faster than any before ever before technology moves faster
43:27
Play from here
Copy citation
than ever before. And so yes, maybe on the other end of this revolution won't be okay
43:32
Play from here
Copy citation
from a societal job standpoint, but the interim, which will probably be like 10 or 15 years,
43:38
Play from here
Copy citation
just because people can't learn new skills quickly. And their behaviors are, especially
43:43
Play from here
Copy citation
as they get older are stuck. They don't adopt and adapt, and we can say, that's not right.
43:46
Play from here
Copy citation
They need to adapt and we can be harsh about it. But the fact is just, that's how the most
43:50
Play from here
Copy citation
of the population is. So I'm, I think there's going to be a lot of job losses. There's going
43:55
Play from here
Copy citation
to be a lot of turmoil in markets and in governments. I think they'll probably.
43:59
Play from here
Copy citation
overreact and try to, you know, equilibrate or redistribute wealth or whatever, however
44:04
Play from here
Copy citation
they go about it.
44:05
Play from here
Copy citation
But yeah, I think there's going to be a lot of disruption on the labor front.
44:08
Play from here
Copy citation
I think most of the conversations I'm having with folks, they're like, oh yeah, we don't
44:13
Play from here
Copy citation
want to cut people.
44:14
Play from here
Copy citation
We just want to make them more efficient.
44:16
Play from here
Copy citation
And I think they're being very genuine about that.
44:18
Play from here
Copy citation
But I think subconsciously, you know, that could change as the technology gets better.
44:24
Play from here
Copy citation
But for now, what's definitely true is they do not want to hire anyone new.
44:28
Play from here
Copy citation
That's the key, right?
44:29
Play from here
Copy citation
Yeah.
44:30
Play from here
Copy citation
Regardless of whether they're cutting or not, it doesn't matter.
44:33
Play from here
Copy citation
But like, they don't want to hire anyone new.
44:34
Play from here
Copy citation
If the company was growing at 20% a year, and therefore its headcount was growing at
44:39
Play from here
Copy citation
if you're embracing efficiencies at 15, 10 to 15% a year, what I am definitely seeing
44:46
Play from here
Copy citation
is it's still growing at 20% a year, but they're capping headcount.
44:50
Play from here
Copy citation
So those extra labor units aren't going to find the homes that they used to.
44:56
Play from here
Copy citation
I think it's a little bit disingenuous for people to say, oh, we're just going to create
45:00
Play from here
Copy citation
all of these new industries, and there's going to be all these entrepreneurial opportunities
45:04
Play from here
Copy citation
when those entrepreneurial opportunities are predicated on leaning into these technologies,
45:10
Play from here
Copy citation
which means that solopreneurs or like partner groups, small groups can build so many more
45:17
Play from here
Copy citation
things without having to hire the administrative load.
45:21
Play from here
Copy citation
So I've done the math and it's, you have to be, if you assume that you're going to
45:27
Play from here
Copy citation
create something like a 50% efficiency in an organization, you would have to basically
45:33
Play from here
Copy citation
double the economy to find homes for all of the people who are dislodged.
45:40
Play from here
Copy citation
If the efficiency is 90%, the math maps out to something like 35%, 35 times GDP of how
45:47
Play from here
Copy citation
much we're going to have to grow GDP in order to create all of those edge opportunities,
45:53
Play from here
Copy citation
be it services, be it new plumbers, like whatever the heck it is for all of those people to find homes.
45:59
Play from here
Copy citation
So I know that you have children.
46:02
Play from here
Copy citation
So how are you talking to them about this?
46:04
Play from here
Copy citation
Well, I'll jump into that one quick reaction and it is, sorry, I do got to go in a little bit here.
46:10
Play from here
Copy citation
I got to text this guy in a second, but a quick reaction is also within our own company.
46:16
Play from here
Copy citation
Like we can build so much more effectively with these tools.
46:20
Play from here
Copy citation
We don't need such a massive workforce.
46:22
Play from here
Copy citation
So just to make this full conversation, full circle, even for us internally, we're struggling
46:28
Play from here
Copy citation
with that because we don't need as many people.
46:30
Play from here
Copy citation
And so that puts some hard decisions on our plate that we have to make in terms of layoffs
46:35
Play from here
Copy citation
and, you know, trimming down the organization size.
46:38
Play from here
Copy citation
So we're at the cutting edge, we're experiencing that.
46:40
Play from here
Copy citation
And it's actually like something I have to deal with as a leader in our company all the time.
46:45
Play from here
Copy citation
Okay.
46:45
Play from here
Copy citation
Now to the fun part with kids.
46:47
Play from here
Copy citation
Okay.
46:48
Play from here
Copy citation
I think the future is very bright for children.
46:51
Play from here
Copy citation
Like, you know, I think part of me is worried, of course, that there won't be jobs for them.
46:55
Play from here
Copy citation
But I, this is where I think the other side is.
46:58
Play from here
Copy citation
I'm more worried about folks that are, you know, 25 to 45, somewhere around there, that generation.
47:04
Play from here
Copy citation
But I think for children that are maybe 10, they're going to grow up using these tools.
47:08
Play from here
Copy citation
They're going to, there's going to be a whole new creator economy and entrepreneurial economy.
47:13
Play from here
Copy citation
And if you can try to encourage your children to be creative and resourceful and have some hustle,
47:20
Play from here
Copy citation
I think the future could be really bright for them.
47:22
Play from here
Copy citation
So my own son, he's going to love this, Kevin.
47:24
Play from here
Copy citation
We, over Christmas, my son loves 3D printing.
47:27
Play from here
Copy citation
I think you love 3D printing, is that right?
47:28
Play from here
Copy citation
I do.
47:29
Play from here
Copy citation
I do.
47:29
Play from here
Copy citation
I'm a 3D printing Botafy geek.
47:31
Play from here
Copy citation
Okay, cool.
47:32
Play from here
Copy citation
Nice.
47:33
Play from here
Copy citation
Awesome.
47:33
Play from here
Copy citation
I love it.
47:34
Play from here
Copy citation
I, I have, I wish I had some stuff here.
47:36
Play from here
Copy citation
I have so many 3D printed toys throughout the house.
47:38
Play from here
Copy citation
I actually was using printers to print printers.
47:40
Play from here
Copy citation
So at one point it came full circle.
47:43
Play from here
Copy citation
But my son's got into it and the tech is so good now for children.
47:46
Play from here
Copy citation
He wanted to sell 3D printed Christmas ornaments throughout the Christmas season.
47:50
Play from here
Copy citation
So we used Lovable and it has like a voice mode, right?
47:54
Play from here
Copy citation
My son's only 10.
47:54
Play from here
Copy citation
He can't like really type prompts, but he can just tell it like what he wants.
47:58
Play from here
Copy citation
He's like, Hey, build me a website that does blah, blah, blah about selling 3D printed Christmas ornaments.
48:03
Play from here
Copy citation
Whips it up.
48:04
Play from here
Copy citation
You know, I'm there helping him out a little bit, but like for the most part, he's able to get this thing up and running.
48:09
Play from here
Copy citation
Then he put some, uh, he prints some QR codes and he goes around the neighborhood and stuff in mailboxes, the QR codes.
48:16
Play from here
Copy citation
And he, guess what?
48:16
Play from here
Copy citation
You want to know how much he actually sold during the holiday?
48:18
Play from here
Copy citation
No, go ahead.
48:20
Play from here
Copy citation
$500, almost, almost $500 for a 10 year old man.
48:25
Play from here
Copy citation
So I always tell people, this is the new age lemonade stand.
48:28
Play from here
Copy citation
Okay.
48:28
Play from here
Copy citation
Don't do a lemonade stand anymore.
48:30
Play from here
Copy citation
If your dad or mom out there, you know, have them build a quick Lovable site and try selling something through Lovable with QR codes at their school.
48:36
Play from here
Copy citation
Isn't it?
48:37
Play from here
Copy citation
Isn't it cool?
48:38
Play from here
Copy citation
I know you have to go, but it's, it's this idea that there's these barriers to entry to various things that are out there and they've, they're just so low right now.
48:45
Play from here
Copy citation
Like, you know, I'm, I, I'm not a coder.
48:48
Play from here
Copy citation
Like I could never build things without using like low code, no code tools.
48:51
Play from here
Copy citation
Now I can build whatever I want, which is absolutely amazing.
48:55
Play from here
Copy citation
And I don't think that most people have really figured that out yet, that if you can dream it, if you have a product story that you want to tell, now you can build that product in a way that, that you just couldn't before.
49:08
Play from here
Copy citation
You don't need to wait.
49:09
Play from here
Copy citation
You don't need to go to university and do a whole coding degree.
49:13
Play from here
Copy citation
No, you just have to have the idea and some wherewithal to muscle your way through.
49:18
Play from here
Copy citation
And then the rest is distribution, right?
49:20
Play from here
Copy citation
Yep.
49:20
Play from here
Copy citation
Now I'm super, it's, it's really cool.
49:24
Play from here
Copy citation
They took something, it's scary at the same time, Kevin.
49:26
Play from here
Copy citation
So that's all sounds amazing.
49:28
Play from here
Copy citation
But think about this, like that used to be a service that we would charge, you know, five to $10,000 for.
49:33
Play from here
Copy citation
Now my nine-year-old son on a weekend can build this website, right?
49:37
Play from here
Copy citation
Like, so at some point, you know, like, Oh, it's like awesome.
49:41
Play from here
Copy citation
Like I see that side, but then depending on the day I wake up, how I feel in the morning, it could be also pretty scary.
49:46
Play from here
Copy citation
Right?
49:47
Play from here
Copy citation
Well, you know, hey Matt, thank you so much for joining us.
49:50
Play from here
Copy citation
This has been super fun and it's, it's cool to hear what you're doing and what your clients are doing.
49:56
Play from here
Copy citation
I guess if I, if I had one parting question, it would be what is your personal tech stack?
50:01
Play from here
Copy citation
Not company, not corporate, but what are the tools that you couldn't live without?
50:07
Play from here
Copy citation
We're using OpenClaw a lot right now and CloudCowork, Cursor, Lovable, a lot.
50:13
Play from here
Copy citation
I would say those are the ones that are always at my fingertips.
50:16
Play from here
Copy citation
All right.
50:17
Play from here
Copy citation
All right, Matt.
50:18
Play from here
Copy citation
Again, thanks for joining us.
50:19
Play from here
Copy citation
If anybody wants to reach you, what's the best way of doing so?
50:22
Play from here
Copy citation
Yeah.
50:22
Play from here
Copy citation
Hit me up at mgram at rapiddevelopers.com.
50:27
Play from here
Copy citation
Feel free to hit me up on X or on LinkedIn as well.
50:30
Play from here
Copy citation
Yeah.
50:30
Play from here
Copy citation
Those are the best channels.
50:31
Play from here
Copy citation
All right.
50:32
Play from here
Copy citation
Well, again, thank you so much.
50:34
Play from here
Copy citation
And we'll talk soon.
50:35
Play from here
Copy citation
Yeah.
50:35
Play from here
Copy citation
Thanks, Kevin.
50:36
Play from here
Copy citation
See ya.
50:37
Play from here
Copy citation
Thanks for listening.
50:42
Play from here
Copy citation
If you'd like to send us a question to answer on the show, email me at Elijah at
50:46
Play from here
Copy citation
spark6.com that is S P A R K the number six dot com or Kevin at Kevin at Ascend
50:53
Play from here
Copy citation
labs.ai to learn more about AI strategy and training for your business.
50:58
Play from here
Copy citation
Visit Kevin's company, Ascend AI labs at ascend labs.ai and to learn more about
51:05
Play from here
Copy citation
design and development of custom AI solutions, visit my agency at spark6.com.
51:11
Play from here
Copy citation
You can also sign up for my free newsletter, AI Framework Friday by visiting
51:15
Play from here
Copy citation
spark6.com forward slash newsletter.