Scaling AI: How AI Can 10X Your Business

If AI is like rocket fuel for your professional service company, how can you use AI to 10x your business?

And are you set up to measure how much faster your business operates thanks to AI?

And what does it look like when projects that used to take three months, now take three days?

Hi, I’m Courtney Baker, and this is AI Knowhow from Knownwell, helping you re-imagine your business in the AI era.

This is the third episode where we’re focused on scaling your business with AI.

As always, I’m joined by Knownwell CEO David DeWolf, Chief Product and Technology Officer Mohan Rao, and NordLight CEO Pete Buer.

To kick things off, I’m excited to bring you a discussion with Matt Pasienski.

Matt is the CEO and co-founder of PACT, and he sat down with Pete Buer recently to talk about how AI is primed to move closer to the executive suite in 2025.

This is a really interesting conversation.

It covers AI management, co-pilots.

They talk about what working in a warehouse in the 1950s can teach us about AI in the 2020s, and what variability in your value chain means.

Matt, thank you so much for joining us.

Thanks, Pete.

Glad to be here.

Just to get us rolling, could you give a little bit of background on PACT and your role in the link to AI?

Yeah.

The way I ended up, PACT is a company that’s building a co-pilot for project management in particular.

The way I got here is I was a services CEO.

I founded a small services company, grew to over 1,000 people.

Used to be a data scientist before that, and I actually have a PhD in physics.

So combining together all of my interests was, hey, how do we do project management now that we have AI and things are changing so quickly, and that’s PACT.

Cool.

With all that education, I’m amazed you had time for getting around to business, but here we are.

It’s amazing how you can express your intellectual curiosity these days and how fun business can be.

But who would have thought?

What a great combination.

Copilots are the focus for today’s discussion.

I was hoping to start there at a relatively high level.

As we think about application of technology to business problems, where does a copilot best serve the need?

I tend to, and this isn’t standard, but I don’t think a lot of people understand how AI is going to be adopted into existing businesses.

There’s a lot of people doing demos and things.

I break it up into three levels.

That three levels is essentially how aggressive or how much agency that AI has.

The first level is you might have a question, you go to open AI, you open up the chat window, you dump in a couple of documents, you ask it some questions.

That’s level one.

I think a lot of people are there.

Level two though is where we get to copilots, which is when copilots in general, you have copilots for writing your emails, for writing code.

For management copilot, for me, that’s when the AI starts to do things on your behalf ahead of time.

It starts to be aggressive, it starts to read things and go and do work on your behalf, but it checks in with you before it does anything big.

And that’s the second level of agency.

It starts to do things, but it’s not going to take any big decisions about resource allocation or anything like that, or telling people what to do until it checks in with you.

And so, that’s where PACT is, and we have a lot of systems that we have tested in real companies with real projects that kind of handle that interaction between human and machine, and that’s what’s super important for co-pilots.

But then the next thing that’s coming, and I think in 2024, 12% of investment up from zero was in agents, and next year it looks like to be the year of agents in 2025, and that’s why I talk about agency, is AI going and acting as its own employee.

And I think that’s where the AI is making resource allocation decisions on your behalf.

So what a terrifically simple and feels right frame, right?

Like there’s the sort of responsive, supportive and independent actor, agents of AI, that’s amazing.

In order to kind of ground our listeners, leaders of services businesses, as you know, what are a couple of the best co-pilot examples you’ve seen?

The biggest one, if I’m leading a services company, is looking at the impact on the individual contributor.

That’s certainly already mature in the sense that every company should be adopting, for instance, software, if you’re writing software, co-pilot.

That’s what I used to do, is we had a software development company.

And so from making sure that you have the best tools, things like the ones that I hear very frequently, cursor is very big, but you have Microsoft and GitHub co-pilot and all those things.

Those are extraordinarily important and should be adopted immediately.

And it makes a real impact in the…

You can immediately speed up projects, but it takes real work to educate your team and to make sure that they feel empowered to take a leap.

And I talk about leap because it’s not something where you’re going to totally be in control.

You need to think about maybe building your company and building building software, or even if you have other kinds of services.

I know my wife works in financial services and her CEO was doing financial services.

And that’s why you’re reading gigantic documents and extracting information.

And I would say that for those types of services firms, it might be even more important because of the capability of ingesting large amounts of information.

So those are the two places I’ve seen individual co-pilots.

And so just to kind of tack on, like I’m here talking about management co-pilot, which I don’t think is a category that people have experienced it.

And certainly not widespread.

I don’t think it’s the essential thing.

Adopt the first ones, which is speed up your individual work.

But what’s coming is then how do you manage that work, which is one of the key challenges as people adopt these things.

You know, I’ve talked to CTOs where projects that used to take, hey, maybe this will take three months.

They have people in there pulling out these co-pilots, getting it done in two or three days.

And that’s going to create enormous variability in your value chains.

And I know you guys think about things in that fashion.

The individual work all the way to the client.

Well, now you have all this variability.

Sometimes it works, sometimes it doesn’t.

And a lot of times you’re not in full control over the outcomes that are being driven by AI when you start using co-pilots.

Maybe it’s in the house style.

Maybe there’s little errors and mistakes.

And so you’re going to have to think about how you control the quality, the speed, the risk in project development and project management in a much different way.

So that’s my take on the industry in 2024 and it’s changing fast.

That’s exciting.

I think it’s easy to relate to the notion of a co-pilot at the individual level.

Can you tell us more what the work of a co-pilot at the management level looks like?

Yeah, let me tell a story first.

I used to be a CEO of a service company.

And one of the things that was very difficult for me when managing interaction with customer, for instance, was out of band communication.

And now the reasoning is simple.

I have an engineer who’s working for a customer.

I might have 30 or 40 engineers on a project.

And so it makes sense.

I have the engineer, they open up Slack, we’re all on the same channel, or maybe have a DM with the client.

Now, what happens though is you get a change in the scope of the overall project is occurring in these channels in a completely uncontrolled way.

So that’s just one example.

And the impact for us is that those types of things, if you didn’t catch them, scope changes or risk that’s in those DMs, you get to the end of the project and the client says, okay, where is it?

And then you learn at the last minute that there’s a different expectation of scope, or that the project hasn’t gotten to the expected outcome because other work has been added.

And in certain cases where we had really big clients that were very important to us, we lost hundreds of thousands of dollars in chunks, not over the whole time, but at a time, because we had to go in and basically make things right, because we hadn’t kept expectations.

So that’s just one example.

And I think we can have a broader discussion about what is management, and that’s going to be a big, it’s going to be a changing, changing topic.

A lot of people are going to be very surprised about how their job changes when AI comes in.

But as a first example of the type of thing that an AI copilot can do, is it can listen to every single conversation.

And that doesn’t just include text, but of course, AI has also enabled us to have perfect transcripts and meeting summaries and action items from every single call.

And so when I’m talking with customers, they’re interested in saying, okay, I have all these sideband conversations, I have project managers, all these different people, the individuals, the engineers, I’m going to take your own services company and you replace those people that are having those conversations.

But not only can I now control scope change, when I hear risk, you know, people will say, hey, this looks like it’s not going to happen.

That should immediately be escalated.

That’s one aspect, but then there’s a huge upside as well.

If you can listen to all these conversations and make sense of it, which is exactly what AI is offering right now, you can do things like detecting new opportunities.

Every time, if you have 30, 40 people talking with customers all the time on these bigger projects, they’re talking about what’s happening at their company.

But a lot of times the people who have the agency to go initiate a new sales process, a new opportunity, they don’t see any of that.

And so that’s, I think, where what I consider the scope of a management co-pilot, it’s super exciting.

Each person should have their own management co-pilot that’s helping them understand and make sense of all of this massive amount of information, which generally is just completely wasted.

I know sales for Slack, you have to pay them a whole bunch of money to save all this information for you.

But I know that the number of companies I’ve seen make effective use of all of those internal and external conversations that are met in Slack is like zero.

And so that’s an immediate place for AI to assist, essentially what I consider the function of management, which is making sense of all these things and looking for new business.

Hard-nosed CFO and I describe the use of AI in my business as the waving of a magic wand.

People think this technology is going to do everything.

If you had to kind of give the line item list in priority of the hard benefits of, for instance, a management copilot like we’ve just described, you gave some examples that I can feather in as you’re talking about.

How would you lay it out?

So, as a fiduciary of a services company, you’re a CEO, you’re a CFO, you certainly have to worry about the long-term value of your company if you are not as productive.

I think that’s probably the first place I would start.

You can look at your internal costs, but we all know that as a services company, that’s not going to be a huge component of your spend.

It is going to be the prices you pay for the most talented people.

You talked about these unicorn people that you’ve paid so much.

You need to make it the maximum output from them.

That’s number one.

And that’s probably the only consideration this year is like I said, adopting AI.

Second, I would say you need to start changing your business practices to account for the fact that some things are going to be 10 or 100x faster than they were three years ago.

And while that sounds like free money, most people, most companies are not set up to benefit from it.

You know, you think about how many projects are time and material.

Well, guess what?

If you start working faster, who benefits?

I mean, the customer will get more money, but do you have the way to show the customer that you’re generating 10x faster results?

Are you building a reputation as having that?

Are you able to account for the fact that you’re going faster?

Are you able to generate more revenue and a decisive advantage against your rivals who are adopting AI slower?

Are you understanding all of those parts of your industry that are most applicable for AI and moving more of your business into those to make additional profits, right?

That requires a lot of understanding that just really doesn’t exist right now.

There’s no data on that.

No one thinks that way.

And so you really need to start thinking in terms of outcome and not the amount of work that it’s going to take because something that took 100 hours might take two.

It might or it might take 100 hours.

I don’t know.

That’s the thing.

You need to start tracking work in a way that will allow you to find these parts of your business that are substantially more valuable.

So I think, again, first one, just go faster with the team you have.

Don’t worry about it.

It’ll work itself out for the next two, three, four, five months.

But as you start to see the benefits, you then need to think about how you structure your relationships with your customer, your employees very differently to account for the fact that certain things are going to work way, way better and certain things won’t.

And you don’t know really which is which.

The logic is pretty clear and you paint a beautiful vision that I can commit to.

I bet there’s resistance in the organization.

What kinds of change management challenges do companies need to overcome?

Yeah, well, it turns out some people don’t like accountability.

And with a very direct measurement of work outcomes, we sit here utopian, hey, yeah, everyone’s going to feel great about this.

We’re going to do more.

Our clients are going to benefit.

We’re going to be more profitable.

But certainly, I go talk to companies and I say, hey, I’m going to make your management more effective.

And the first thing that goes into a lot of people, you have mortgages and kids and things like that.

Maybe it’ll be too effective.

Maybe I won’t be.

And I think the same goes for software engineers.

You go on a message board with the best and brightest that are coming out of school.

They’re all concerned about their career because now here’s something that does their job for them.

And that’s real.

I think the question for companies is, can you create the social circumstances so that you can adopt this stuff and become productive with it, just like we did with cloud or software in the first place?

Certain businesses survive because they were able to adapt and they had a culture of trust that allowed them to do that.

And then other ones waited for a new entrant.

But I guarantee you, any services company that’s doing $50, $100 million of business, there is a venture capitalist who knows how much you make and is funding a 25-year-old to come and replace you.

So, you know, I don’t think that’s going to succeed in every case, but you need to be concerned about it.

And it is a, you need to fight that pushback that you talk about because it’s there.

It’s absolutely there.

It’s going to be there.

I feel like I understand academically the directive to build trust and create the social circumstances that enable working in a different way.

What does that really look like in practice?

If you’re measuring your work on another level, if you are using data to run your company, and you’re worried about change management, or you’re worried about retaining your incredibly good engine, think about the best 20% of your company.

Make sure that that data makes them look good.

When I sell the clients, everybody gets it when I say, our job is not to make the company successful, it’s to get our buyer promoted.

If we do that, everything else takes care of itself.

But the same goes internally.

My job when someone joins my company, if I’m a services CEO, is to make that person, if they come and kick behind, like they’re gonna have a incredible career by working at this company.

I’m gonna get them promoted.

I’m going to make sure that if they’re engaging with the high-tech, if they’re engaging with AI, if they’re pushing…

I have a story from my dad.

He got out of the Army, and before he started his real career, he had a few months where he was working at a warehouse.

And he came in, he came out of the Army, moving boxes, shipping stuff, and some of the guys came to him, and they said, hey man, Pasienski, stop it.

You’re going to blow all of our quota.

You’re going to blow our quota.

Stop it.

Slow down.

And, you know, that’s, it’s a warehouse.

What it was, and he left a couple of months later, but what he did is he would just pull his tickets and then go take a nap in the back while he was studying at night.

So it helped him out.

But like, my point is, think about the culture of your company and how are you creating the circumstances when people do come in and do 10x faster work.

Can you measure it?

And can they benefit?

Think about how beneficial it is to you and your customers.

Can they benefit at least some of that?

And most people don’t think about it like that.

They just do it the old way.

And I think that’s going to be a big challenge as variation and output really becomes huge with, at least in certain parts of your industry.

In a way, well, I guess what we’re talking about is leaders adapting their talent, inspiration and development methods, right?

It has always been true that the best leaders do the job of painting the vision of what a great future can look like.

The difference now is, and you will be powered by a technology that will make you better, stronger, faster, and may take some of your work away, but then there’s other things that you can be doing that are incredibly value creating, right?

Paint the picture of the upside for everybody.

You know that if you could do twice the work for half the price, you could go sell more.

Just make sure that your institutions, somebody is looking at it.

And I think the double edged sword of there’s going to be resistance to using data can also be incredibly powerful, that you could retain your very best people and retain your customers at a much higher level.

I think that’s the magic outcome for everybody, but it requires engagement with the problem.

It’s not just do things the old way with data, with AI, and hope it works out.

It’s like, that’s why we see so much.

I mean, there’s a lot of things people could be doing now five years ago that they weren’t doing for the same reason.

So it’s always a challenge and it takes great leadership.

I think that’s a wonderful place to end, Matt.

It’s been a pleasure.

Thank you so much for sharing your thinking and for joining in today.

Hey, love what you’re doing, Pete.

Always appreciate talking with you guys, and thanks for inviting me on.

Are you tired of losing customers, especially the ones that you’re surprised about, that you didn’t see coming?

We all hate that.

And that’s exactly why we built Knownwell for professional service firms.

If that’s you, and you’re tired of being surprised by churn, go to knownwell.com to learn more and sign up for our beta wait list.

After listening to Matt’s interview with Pete, I was excited to get in the studio with David and Mohan to get their take.

David and Mohan, we’ve talked quite a bit on this podcast about the five levels of work.

And he really, it sounds like I might be a fan of that work or those things that we’ve talked about, specifically on the operations side of things.

What else would you add to this idea?

I think he gave some really interesting use cases.

You know, the first thing it did for me, Courtney, is it brought me back the last conversation we had.

I think it was Charlene Lee was on the podcast and was talking about replacing the CEO, right?

And AI.

And we talked about this idea of co-pilot and the more complex, the better off the AI is to actually help us with those decisions.

And really, that is what Matt is talking about, but at all layers of the organization.

And I love the distinction, right?

He talks about the management co-pilot.

And I think there’s two ways to look at that, right?

There’s the management of, just as we’re thinking through, driving this knowledge work of execution, right?

The accountability, the follow up.

And then there’s the operational component, which is kind of connecting the dots, right?

If you think about the example he uses on listening to every single conversation, right?

Listening to every single conversation and getting a recap of the meaning and the transcript and the follow up items, it’s very execution oriented, right?

It’s doing work.

It’s processing.

It’s intelligent.

It can even hold you accountable of, hey, you haven’t followed up on this yet, right?

That’s a management co-pilot, right?

But if you can start to look across calls and identify trends and start to get into what he was saying about identifying opportunities for growth within accounts, right?

Those types of things.

Now you’re talking about the operations of not just doing more work, doing it more efficiently, which is the management piece, but the operations piece of systematizing that and figuring out how do you do the right kinds of work and make it more efficient holistically as you pull these individual processes together.

And then the second thing I think about is, oh, isn’t that funny?

He happens to love Knownwell, apparently.

That’s what we’re building.

You know, when Matt was talking about it, I was thinking something very similar to what David said, but I’ll put it in two buckets.

That is the efficiency bucket and the effectiveness bucket.

Right?

So there is the efficiency bucket of saying, how do I consume large amounts of information that I don’t even have access to because somebody is talking to somebody at your customer or that you have no idea that they’re talking to, but if you can get that information and get that summarized and get that properly kind of shown to you at the right times in the right context, right?

That’s all about efficiency.

Like you would have otherwise be chasing and slacking people to get that information and then it might not even be perfect because it’s after the fact.

As opposed to effectiveness, where what if you could make eight or nine out of ten decisions right and get it correct in retrospect, as opposed to five out of ten, right?

Just because you had the ability to roll the tape forward and see the future a bit and get to these micro predictions and then come back and make good decisions.

So, yeah, absolutely.

Everybody deserves a digital chief of staff.

That’s how we think of it.

And there are two components here between efficiency and effectiveness.

I love that story that Matt tells about his dad getting too much work done.

And honestly, I could kind of see that happening today as, you know, certain employees are deploying AI and other ones aren’t.

What are your thoughts on that?

You know, I think it’s such a brilliant point that Matt’s made, right?

So one of the things that we know in any business is that you have to look at it as a value chain of processes, right?

When you optimize for one particular process, new constraints are going to emerge on other sides.

You can think of it as software development, you can think of it as warehouse operations.

Maybe you can kind of get things into the box well, but maybe your shipping carrier can take it at the same rate, right?

So new constraints are going to emerge and it’s going to have massive cultural impacts as these constraints emerge.

And I think it behooves leadership to get ahead of it and look at this as an end-to-end enterprise when you’re applying AI.

And when you get to 10X in some parts, what does it do to other parts of the organization?

I think there’s two other aspects of that, right?

The first one that it prompts me, go back to the idea of management and CEO co-pilot, like growing the business together and not playing whack-a-mole is one of the biggest challenges of being a CEO, right?

What you just described is something AI can actually help with, right?

The other piece that I think he so keenly brings up in that story is the impact on other people.

I think as we’re going through this AI transformation, it’s so important that we don’t leave people behind.

There will be some people that embrace it, there will be some people that are excited about it, there will be some of these people that are scared about it, there are some people that abuse it and use it to be lazy.

There’s all sorts of different types of people.

And I think not forgetting the human component of the change we’re going through.

Again, go back to the industrial revolution that we’ve brought up as kind of an analogy we can look at.

It transformed people’s lives.

And we often look at just the job loss part of that, right?

And whatever your perspective on that, there will be job disruption.

But I think you have to look even further than just jobs.

You have to look at the transformation of the human experience if we really want to create a world that is better off because of AI.

And so I think that component is just as important.

So as we wrap up this conversation, talking about how AI can help you connect your business, and specifically this management layer, this operation layer, and how AI can be used.

Any final thoughts as we wrap up this conversation?

You know, Courtney, one of the things that just popped into my head as Matt was talking there is, yeah, we’ve talked a lot in this session around the management and the operations.

I couldn’t agree with him more.

That’s a 10x.

I also think that one of the 100x plays that’s out there that’s not very far behind and is already being used actually to some degree is the ability to do strategic work as well.

You know, making strategic decisions is oftentimes fueled by data and a lot of rationalization and the type of analytical creative work that AI can actually really help us with.

And I would look at that next horizon too.

It is not just the management and the operations.

It’s also the strategic and how do you deploy AI to make sure it’s helping you differentiate further and understand your market position and those types of things.

There’s already some of that out there, but don’t forget that next step either.

The variability and mismatches are going to be there.

So if you look at how architecture is coming around for agentic AI, there is something called an agent core that sits in the middle, right?

So whether that happens through a computer doing it or a machine doing it or in the short term, the next few years, a human doing it, you’ve got to have somebody who sits in the middle and is able to kind of adjust for these variabilities.

Maybe one side goes faster, inevitable in the short term, the other side won’t be going faster.

And that middle piece is called good leadership, right?

And be able to manage all of these mismatches and understand how to compensate for it.

In agentic AI, we call it agent core.

Short term, it’s good management.

Pete Buer joins us as always to break down the business impact of some of the latest and greatest in AI news.

Hey, Pete, how are you?

I’m good, Courtney.

How are you doing?

I’m doing well.

This week’s story brings us to the intersection of art and AI.

The New York Times recently covered the humanoid robot, Ada, selling a painting at auction for $1.1 million.

What do you make of this one, Pete?

Well, first of all, I did a little research.

Thanks to Chet GPT, I discovered that roughly 100 or so paintings in the US each year sell for $1 million or more.

I think more than anything, Ada’s success challenges our assumptions on where the line gets drawn between human and machine on creative endeavors.

We had kind of up until now, or at least I had assumed for the most part, that the roads, the mechanical gets done by one, and the creative and unusual gets done by the other.

But not the case.

The creative now we see can just as easily be done by AI.

It’s not just execution given a design, right?

It was conception and representation.

Ada called the shots on what this would be, and what it would look like.

So summing it all up for me, this is just another reminder of the world of possibilities that AI opens up for us all.

We’re making new discoveries every day, it seems, on the next new thing that AI can help us get done.

I think the implication for us as leaders has to be to take a look at the work, the job, the society we’re trying to improve and ask ourselves daily whether we’re getting the most from AI in our pursuits.

I love that.

So interesting.

And I do, you know, this is an audio podcast.

Although if you’re on YouTube, we can help you out with this.

But I do feel like for everybody out there, do you want to just describe the painting a bit just in case they haven’t seen it?

No.

I have to be holder, right?

I can’t lend any credible take to what looks to me a little bit like somebody used the sponge to outline a person’s face.

That’s kind of as good as I can do.

I do feel like it has a little bit of ET vibes personally.

Whether subliminally or purposefully, that would be pretty funny.

Yeah, for sure.

Well, Pete, thank you as always.

Thank you, Courtney.

Thanks as always for listening and watching.

Hey, we would love it if you would leave as a review.

Seriously, it really helps more people find the show.

At the end of every episode, we like to ask one of our AI friends to weigh in on the topic at hand.

So hey, perplexity, what’s happening?

Welcome back to the show.

This episode, we’re talking about AI strategies and tactics for 2025.

So what do you recommend?

In 2025, AI strategies will pivot towards hyper-personalized experiences with robust privacy protections.

The key is smart contextual interactions that feel tailored yet respectful of user boundaries.

Now, you’re in the know.

Thanks as always for listening.

We’ll see you next week with more AI applications, discussions, and experts.