Turning AI From Hype Into Action with ForwardPath AI’s Josiah Shelley

In this Snippets episode, Leon speaks with Josiah Shelley, CEO and Co-Founder of ForwardPath AI, about how organizations can move from AI curiosity to practical implementation and measurable results.

Josiah shares insights from working with hundreds of businesses across industries, highlighting the most common challenges leaders face when adopting AI, including over-reliance on AI committees, lack of hands-on experimentation, and uncertainty around where to begin.

He emphasizes that successful adoption starts with leadership engagement, encouraging executives to actively use the tools themselves before scaling across their organizations. From Copilot and Claude to custom-built agents, Josiah discusses how real value comes from understanding and applying these tools in day-to-day workflows.

The conversation also explores how AI is reshaping workforce strategy, from automation and efficiency gains to new opportunities for upskilling, revenue generation, and rethinking organizational structure.

Tune in for practical advice on moving past the hype and building AI strategies that actually work in practice.

Transcript

Special thanks to Doanne Grant Thornton for helping us bring you today's PEO Leadership Innovators Alliance Snippets Podcast.

00:13
Welcome to our snippets podcast. I'm Leon Gore and CEO, PEO leadership and innovators Alliance,  North America's premier peer to peer networking and leadership advisory firm.  Today, I'm very excited to welcome Josiah Shelley, CEO and co-founder of ForwardPath AI and member of our leadership community since 2020. Josiah is a serial entrepreneur who's founded three ventures and successfully exited two. He sold a mobile marketing SaaS company back in 2012, long before mobile first was table stakes.

00:43
And most recently in 2020, Fourier Exeter Marketing Agency, where over five years he helped the business double revenue into eight figures and quadruple profits. Boy, I hope I'm right there. I'm reading this here for you.  he's an operator who's built scaled and sole companies, and he brings that discipline to how he thinks about technology. Forward Path AI is a Canadian practical AI firm that helps mid-size organization students move from AI curiosity to deploy tools, train teams, and measurable outcomes.

01:13
I remember sitting with Josiah, maybe it was over two years ago now, Josiah, we're sitting in a coffee shop. You were sort of talking to me a little bit about the idea you had and the venture and we should be moving into AI as an organization as well. I was like, no, no, no, no, no. And now I look where you're at today and I'm like, holy smoke, this thing has just launched. It has been fantastic. So welcome to the podcast. Thank you. Thank you very much. Yeah. Everybody's trying to figure out what to do with this AI thing. So that's played well in our favor.

01:43
Well, I'm going to, I know members are going to listen to this and they're all sort of battling through different things. So let me sort of kind of cut through the hype in our leadership community. What we're finding is people are being pitched all the time now in AI. They got to get it within their organizations. Some of them have still not moved on it. I'm just curious. You are seeing a lot. You've been there at the very beginning. Where should they actually go? Where do they spend their first dollars and how do they make that decision in terms of?

02:12
where to go. They know they need to go now into that. What I've observed is that AI can get implemented in your organization only to the speed that the leader  is willing to lean into AI. Like  if the leader is willing to test and work with AI and figure out how it's working, how it's not, and educating themselves,  you can't  move faster than the leader.  Your risk tolerance, what you're willing to test and learn with AI.

02:42
And so I think when you look at like, where do I put my first dollar? You first need to make sure that  you don't need to spend much money, but you need to first educate yourself on what AI is and what it isn't and how it works. You need to make sure your leadership team is aligned to why even embark on this AI journey and what narrative and story are you going to share in the organization? So people understand why AI is being implemented in the first place.

03:10
Because what you can't do is say, let's lead with technology and we'll buy another tool or  we'll build a tool and then say, everybody needs to use it. Like, why aren't you using AI? Why aren't you using Co-Pilot or why aren't you using Claude? You kind of need to understand like, what does this do for us? Be on board with it at a leadership level and then that will cascade and help the organization. And so like, these are the things like I'm observing.

03:36
After going through custom builds and building these tools and deploying them, there are, it's a lot more than just technology  discussion, change management, the right people in the right seats,  and then really C-suite that are aligned to why they're doing it in the first place and where they're going with it. agree with you. Educate yourself first, right? And we're sort of pushing our members out and then every group we're sort of saying, okay, where are you at now? What have you done from last month to this month? We're still getting the stragglers there.

04:06
And the thing is moving so quickly. Like even myself, I think about it. You had ChatGPT, had OpenClaw, which is  crazy and then came back. And then you got Claude and Claude keeps releasing new releases. It's almost, we're asking these leaders to educate themselves, but I don't even know how they educate themselves because it's constantly changing so fast.  good place to start is just pick the tool that's in front of you. So if you're using ChatGPT,

04:36
or maybe you're using Claude's or Co-Pilot. Maybe your organization has already got some licenses they've already subscribed to or paying for. Like first,  educate yourself by just using the tools and using the tech. You're not gonna learn by reading another article about how some new model came out and it's gonna change the way design is being done. You're gonna learn by actually playing with the tool and getting good at understanding how the tools work.

05:03
Because I think once you have your aha moment where you're kind of like, wow, this thing is powerful, like more powerful than I thought it was. And it did something that I'm surprised it was able to do. When you start having those moments, that's how you turn a straggler into somebody who's bought in and now going to start using these tools. And so I would,  I would encourage like  really play with tools and expand their capabilities and figure out how to get something done with them. They get past that stage.

05:33
Now they bring something back into the team. What have you seen in terms of successful teams or CEOs or presidents or the exact getting their team aligned? Who's done it really well and how they done it? When I've seen it done really well, the CEO  is not punting it down to someone else to figure out. There's a tendency to create an AI committee. So you say, you know, I don't really want to be involved, but I'm sure somebody else can like run this thing and figure out AI for us.

06:02
Let's build an AI committee that operates for six, eight, 12 months. And if you're critical about it, you say, what have you actually done? Like not just talked about and what have you actually done AI committee often after eight to 12 months, all they've really done is a policy and the policy needs to change because the tech changed so fast. Like that doesn't work. This idea of doing it off the side of your desk and then giving it to somebody else to do,  and they have their day job and they're trying to figure out AI. I don't see that working.

06:33
What I see kind of working really well is a CEO who is a part of the discussion, like a part of the committee, a part of actually leading the change, using the tools, sharing with their teams the new things they're learning. You have to get your hands dirty again. Like you can't let the computer kind of come out and then say, well, the computer is going to change everything, but I'm not going to use it. And I'm just going to like let my team figure out where a computer is going to be used in the org.

07:01
This in the same way is a big technological advancement. think we have to, us as leaders, as CEOs, we need to use these tools. It's funny,  yesterday, what was it? Tim Cook announced his replacement, right? And you're sitting there going, gee, wonder if he, an engineer is going to take his role. But you look at even our, you know, the current membership and I look at the different leaders, you know, we always talk, you got to keep pushing it, you got to keep pushing the learning, the learning and learning. It's a lot.

07:30
You're asking a lot of these leaders to lead. Plus, they still got to carry on their current role and responsibilities in terms of growth, right? I don't know.  Maybe Cook had the right idea. Step aside. Let the younger guys take over.  I think, you know, even  one of your points on how do you have time to do this in the first place, that is why there should be an exercise of automating what you can in your own role, even as a CEO.

08:00
how I run like forward path. have like 11 AI agents running at any time. I have an AI financial controller. I can make certain decisions that maybe other organizations can at their size and scale, but you should, if you've actually done a deep dive and you're using these tools, you can build personal workflows that free up a lot of your time. You can essentially have a executive assistant style support. Every CEO should have one.

08:28
And you can do that with these existing tools while building out some agents to support your workflows.  like there is the, I think it's an excuse to say, I don't have time. I think if you do set aside some time to learn the tool, figure out how you can implement it, that there are quick wins you can get to free up some time in your day to use that for the thinking you need to do.  I just want to touch on those agents, right? Cause I've been doing that too. Yeah. And so you've got, the agents are like almost like certain

08:57
has to be done. You talked about the financial and different vertical.  For you, are they running 24 hours or are you having to go into those agents and almost create a new project for them each time? So for me, I'll use the financial controller as an example. At Ford Path, we have an AI financial controller that has access to both our bank accounts, QuickBooks,  our sales CRM.

09:25
and account planning for each client on what we want to do with them in the future. So now it has  read-only access to all of this  critical data that you can run cashflow projections, cashflow scenarios. uh You can run forecasts against where we think we're going to end up in 12 months.  I have that run three times a week for me and produce a report, a standard report that I  built because I have the context of the business.

09:52
And  I've done this in my historical life where I had a bookkeeper, I had a controller, I had a CFO, and I'm getting  much better insights into my cashflow because I have built the agent and understand the business intimately. And so  that financial controller, I don't let it autonomously run, make decisions, change anything. Right now, stage one of the agent is just producing reports that it does some analysis and pulls out what's happening and scenarios.

10:20
But I review that three times a week and make decisions in the business around that. And so that's an example  of  one that I rely on.  So we're relying very much on your own judgment, right? You're actually reviewing this stuff. It's running, it's making it much more efficient, but ultimately you're still controlling it. The big value add is the  scenarios that it can run because  it has unlimited time to run scenarios against all of our cashflow scenarios and our sales CRM data.

10:49
It's very good at giving me an in-depth analysis, but ultimately  I want to drive the ship. I want my AI agents to support me with data and information and give me more context, but I want  to keep driving. I'm  not fully bought into autonomous AI running 24 seven,  making decisions on behalf of me and representing it. That's not a  realistic, that is a very high level of maturity. I have a few agents that do

11:18
like execute and do work, but most of it is about analysis and giving me information and then helping me in my day. So let's come back in towards your clients and what you're seeing there and also in the marketplace. Are there some easy wins or like big opportunities within organizations today that people should be really looking at?  Like almost no brainers today. The no brainer  is  so I think the unique perspective

11:48
I have is that  now I've had hundreds of, like I've seen what's happening at hundreds of businesses now, these conversations across the board, different industries. The most common scenario  is AI committee, co-pilot licenses. They've bought some co-pilot licenses. They have no idea how people are using co-pilot, if they're using it at all. And just doing  some co-pilot training is a great place to start because they're already paying for these licenses that

12:17
The MSPs have done so well at selling, now getting some quick wins, honestly, it's like train your staff on whatever tool you have. If it's Co-Pilot, if it's ChatGPT, there's more gains to be had there. And what I've observed is there's not a lot happening. And then really challenging this AI committee model that produce results because most companies have some groups, working groups, some AI committee, but they're not getting to use cases where you can actually implement AI and then either like,

12:47
buying tools or building tools to support workflows and use cases in the org. And so they kind of stalled, but I think an easy win is copilot training, lots of good ways to do that. We do that.  Or with really giving your AI committee a boost  to get valid use cases and get those  kind of built or bots for tools for your teams to use. So you're talking about copilot and I was in a meeting the other day and they were talking about copilot, but they're also talking about code way more now.

13:17
Can you use ClouD Team in the same way, essentially as POPILOT, or can you choose one over the other? need both.  ClouD will integrate into Excel or Word. And  ClouD, I would say, one of the, it is probably the highest performing model, but it is not directly integrated into your Microsoft Stack, into your tenant.  So when you're talking about security and all of those aspects, uh

13:45
certainly Copilot is the easy one to say yes to just because it's already built in your own environment versus opening up to Claude, being dependent on that. Claude for sure is our, like it's my daily driver. It's what we use  day in and day out, but you can't ignore kind of the tools that are already integrated. Most people have the Microsoft suite and so they Copilot already.

14:08
It can do some good things.  Some areas it flops quite heavily and they'd be very transparent. I think even the Microsoft team to admit that, but they're definitely working on  improving it. I would just say whatever you have, like get good at using it. uh If you're totally agnostic, you're like,  I just want to use the best tool out there, Josiah, what is it? Claude for sure today is the best tool.  And maybe that's where the CEO starts or the executive starts, especially if they're to do it independently and you want them to learn.

14:38
All right, so you come back to the AI committee as sort of the biggest stumbling block  in many ways. And it's funny, I've heard that too in the groups. It's like, I got this AI committee. go, what did they do in the last four months? Well, not much. They met regularly. And it's no fault of the organization. It's just AI committees are not set up to really be successful when it's uh kind of punted off the side of somebody's desk.

15:06
So it's funny, I'm going to look hindsight, like a year ago, you and I sat, you were at the conference as well. We presented the conference and maybe it was even two years ago. It was all theoretical. Last year at the conference started moving in a number of companies starting to use it. We are today, I would tell you, everyone knows about it. Probably 90 % or more use it personally and are evolving in the different platforms and stuff. It's common, like a freight train.

15:36
What's, I was going to ask you what's three years look down the road, I mean like what's 12 months look down the road within these organizations. So when it comes to strategy and what you do as an organization,  I think the biggest opportunity is for midsize to small orcs where they can move faster, be more nimble. And as the technology advances, they can embrace it  fully because as an example, Ikea automated a lot of their customer service.

16:06
And so I think it was about 8,000 or 9,000 employees were made redundant essentially within their customer service team because they now have AI supporting it. IKEA could have either  cut those people out, laid them off like all the headlines are doing. we're doing all this AI stuff, so let's lay them off. IKEA chose to create a new revenue stream  and they upskilled and re-skilled this group into interior design consultants is what they call it.

16:35
And they said, wow, maybe we can actually make money off these people. And they created an entire new revenue stream for Ikea. They did a billion dollars, this group, this 9,000 people last year. And I think organizations are going to have to make these choices. we cutting head count? Are we creating new revenue opportunities? Because it would be quite sad if the only kind of objective organizations are finding is cost savings.

17:04
I want AI to support the grunt work and support the stuff that's not fulfilling and give people time back in their days. But I think right now organizations can be so creative on what the future looks like for them because this technology is incredible and it can automate and do a lot of actual work. So where are we going and where does this all end up?

17:26
I just want to challenge business leaders to broaden their view outside of just cost cutting measures to reduce headcount and into how do we be creative, open up more up-skill, re-skill opportunities for revenue generation, support the growth of an organization. Because no question margins have been  squeezed over last few years and we have to continue to grow. But these are the types of things I see happening and I think about their

17:54
It is not just one path, which is adopt AI, cut head count and be more profitable. I think there are many different paths AI can take us on. I hope we're ambitious enough to do that. I would be very scared  of  new entrants as competitors that are fully AI native into my industry. Because they will not think the way that we've,  we're thinking about growth and we're thinking about organizing teams and what teams do. They're going to be thinking about

18:22
how do AI tools enable us to grow as fast as possible? And we're seeing this, we're seeing companies that have very low head counts and they're growing like crazy. But it's very interesting to me to think about, if you're launching a new business, new competitor, you have some college students, know, jamming out and figuring out what AI tools they can build and how it's gonna support their business. It'll be very interesting to see how that plays out. Josiah, thank you so much for sharing some of those insights.

18:50
If you're interested in our live webcasts, The Way Forward Live or any other snippets, please take a moment and visit us at peo-leadership.com. You'll find on our site various prerecorded webcasts, which include some well-recognized guests, such as Morgan Housel, Professor Janice Stein, Harvard's Rosabeth Kanter, and of course, some great leaders from our own community sharing their best practice and ideas, such as Mike Sherrard, David King, Dave Douglas, Dave Jones,  Josiah just right now, and the list goes on.  Thank you for joining us today and we look

19:20
forward to seeing you again shortly.