Insight ON The Org Decision Most CFOs Are Too Cautious To Make

Audio transcript:

The Org Decision Most CFOs Are Too Cautious To Make

Speaker 1 (00:02):

Most organizations by now have access to AI tools. Many have someone whose job title includes the word AI, and almost everyone is still waiting for something to actually change. The problem isn't the technology. It's that no one is accountable for the hard work between tool access and transformed workflows.

Speaker 2 (00:21):

Transformation is not simply, oh, ROI. That doesn't mean you've transferred it, transformed anything. And it's also not adoption. Transformation means that the work you were doing is no longer being done in any significant way like it was being done before. A couple months ago, invoices would come in and a team member would look at an invoice and they would code them and put them in the system. And it was a very high touch manual process. What we've been able to do with AI is allow AI to do those analyses, to code the invoices, to understand what an invoice is and what it relates to. And now the accounts payable professionals are just watching the flow of this. How

Speaker 1 (01:07):

Do you know when you're truly in AI transformation mode versus just doing pilots? When

Speaker 2 (01:11):

People stop talking about AI, when we've truly transformed either a team or a team member, they are no longer doing AI, they're just doing their job. And that is truly the metric of transformation and success with AI is where it is just part of how you do business.

Speaker 1 (01:27):

My guest today is Alex Pugh, Director of Finance AI Operations at Insight. Alex was brought in by Insight CFO to help that finance organization move past early AI wins and into actual workflow transformation. A distinction, as you just heard, he takes very seriously. Alex built his career at the intersection of finance operations and intelligent automation, starting in robotic process automation and working through to large language models always inside finance organizations. In this conversation, you'll get why nine out of 10 problems people bring to an AI team are not actually AI problems and what to do with them anyway. How Insight's finance organizations structured a dedicated AI operations team and why their domain expertise proved essential and what accountable AI ownership looks like. Not the AI mascot, but the doer who brings the receipts. This is Insight on AI Transformation. Let's go. Alex, welcome.

Speaker 2 (02:27):

Thanks for having me.

Speaker 1 (02:28):

We're going to talk today about something that a lot of organizations are grappling with, but I want to start by understanding your journey to land at Insight because you were hired essentially by the CFO to do a role that I think some organizations or some people might consider. I don't want to put you down, but you have one of those titles that like, is this real?

Speaker 2 (02:52):

What

Speaker 1 (02:53):

Do you do?

Speaker 2 (02:54):

Yeah. And that's part of how I can be successful here is no one knows what I do. Yeah, so my journey here to Insight started in October, so we're almost at a year. And I was brought in by the CFO and his senior leadership team to help service the mission, which is to make Insights Finance the number one AI finance organization in the Fortune 500. The CFO laid out that mission, James Morgado, back in February of last year. And they self-organized, went down this path of trying to bring AI to bear in the business. And around October started coming to the conclusion that they needed more technical expertise and depth to kind of go 2.0 on that, to kind of turn the wheel and move into the next phase of that AI journey. And so I have a professional experience in both financing and AI.

And when they found me, I thought it was very impressive like, wow, how'd you get to me? But yeah, so they brought me in. My original title was AI innovation advisor. The only one at Insight. They made that title specifically for me so I could kind of execute on the mission. It was a very helpful title because no one else knew what it was or what it is I was doing here. So I was able to actually traverse the organization in a way that was a little more impounded. It helped me get a better idea of how we can execute and be successful with AI. Since then, my title has been more aligned to our kind of traditional hierarchy. So I'm a director of our finance operations, AI operations, which kind of makes it make sense more about what I do here.

Speaker 1 (04:45):

It hearkens back to the time when innovation was the buzzword of the day and you had the leaders of innovation, heads of innovation. And again, it was kind of like, well, what do you do? You're kind of inventing your role here. And the AI piece I think is particularly interesting because AI itself is not new, but generative AI is still relatively new. There's not a lot of quote unquote experts. We're all learning, experimenting, figuring this out as we go. So if I may be so bold, Alex, what makes you the expert to help an organization, especially a finance team, figure out how to make AI work?

Speaker 2 (05:20):

Sure. And that just goes back to my professional experience working in these finance organizations, bringing to bear this kind of intelligent automation, AI field work and technology. Originally starting in RPA, robotic process automation, and then moving through that whole kind of field of study. So classic machine learning and AI operations to where we're at these days with large language models and that generative AI. And having spent a career in both retail finance, service center finance, and now here at Insight and our finance org has given me this real depth of experience in not only the finance use case, but also the technologies we can bring to pair to solve for them or create new transformative ways of working with things like generative AI or other technologies in the field.

Speaker 1 (06:18):

So what I'm hearing is really your background in learning new technologies, rethinking workflows, bringing some of those new technologies to bear, that's the skillset that's transferable to make you the right person for AI initiatives.

Speaker 2 (06:31):

Correct. But it's also my experience in the context of finance. And while not a practitioner of finance, I'm not a CPA, I'm not an accountant, I don't do FP&A, but I've worked in the field. And so it's that familiarity with the finance organization, a finance organization, like use cases, and that deep experience with what that looks like and what the outcomes need to be, matching that with the technology that I've been able to be successful. And I've seen others be successful in this. Yeah,

Speaker 1 (07:05):

You're emphasizing domain expertise.

Speaker 2 (07:07):

That's correct.

Speaker 1 (07:08):

Which then I guess begs the question, as organizations are thinking about how to really successfully do AI transformation, do you hire the head of AI transformation to lead organization-wide? Or is it better off to have somebody who is specialized within each business unit or business function to make sure that, okay, finance is doing this appropriately, marketing's making progress, sales, et cetera. And if you go that route, how do you do that effectively so that you're kind of organizing the chaos that could happen? Especially with agent building. We hear agents sprawl all the time. So what would be your advice there?

Speaker 2 (07:46):

Yeah. And it's going to depend organization to organization. I'm just going to put that caveat right there. And it really depends on how you're measuring success and again, what your organization looks like. For our own example here at Insight and Finance, we went with originally back in February 2026, they sought to have that domain expertise as the real focal point of our AI transformation, which means they sought to find those finance resources, those resources that already existed in accounting, in internal audit, and our AR and AP functions that could then work and upskill on AI so that they could correctly bring it to bear and transform. That's a little different than a more traditional approach or bringing on that external kind of talent or innovation to come and try to bring the technology closer to your business. They're both valid. I think our approach here at Insight is working and did work and was the correct way to go.

And that's a testament to our CFO and our SVP of tax and treasury on identifying that as a way to do it. We also here at Insight do have a centralized AI approach in our AI COE, which we work in tandem with. And so I think it's striking that right balance of what resources are you willing to bring to bear to solve the problem? One thing we try to bring up, and this is something we hear a lot of people say is AI is transformative. It's this technology that's going to transform the way we do business. If that's the case, we need to transform the way we're organizing around it. We can't expect this transformative technology to come in and transform the way we do work the way we do it currently. It has to transform the work. And we should also transform the way we organize ourselves around it.

And so that might be elevating resources that already exist and expecting them to help bring that AI and bring that transformation. But it also could be bringing in that external to literally bring that fresh perspective and that fresh take.

Speaker 1 (10:00):

You mentioned that when they brought you on board, there was already progress happening. The finance team was already rallying around this vision from the CFO to be very AI forward. What was the blocker? What was the justification for then looking for outside help? And I think in our case it was acceptance. We want to bring someone internally, but this might also be the moment in time where an organization looks for consulting or services to help with this. What was it?

Speaker 2 (10:30):

I mean, just going back to the journey that started back in February, they went and the call went out. We're going to be AI first and therefore we need to elevate and identify those resources that already exist in the business and finance to help bring this AI to bear. And so that's where the Thundercats came from. Our AI champions who - You

Speaker 1 (10:52):

Got to explain that. Who

Speaker 2 (10:54):

Were

Speaker 1 (10:54):

Thundercats.

Speaker 2 (10:55):

So the Thundercats is, we're really good at branding at all levels here at Insight. And so the Thundercats were and are the name of our AI operations team. Now at this point, they were full-time resources dedicated to this effort, but they were self-organized and loosely done. And they called themselves the Thundercats and were called the Thundercats. And that branding has kind of also helped them in their mission and served them well. But it was that allocating and making this their full-time job of using AI and helping with AI adoption that really got the momentum going. There was a point where, and this is when they started looking for the external resource to bring in and help them take it to that 2.0, is they found that they were able to really get the low hanging fruit, the quick wins. They were able to identify the technologies that the business could use.

But at a certain point there was a scale or a technological barrier that they were having trouble getting around. And that's the point they knew, okay, we need to go and find and engage with an external resource, whether that's an external partner or whether that's bringing someone in to kind of help lead that next step. That's what they were able to identify. And I think all orgs should be very capable of identifying when that moment happens. It's just making sure to listen to your resources that are telling you, "Hey, we're just not able to turn the wheel here. We're running up against something and we don't know what it is. And we need resourcing for that." And that will come up from your resources internally if you listen to them.

Speaker 1 (12:40):

I love that they're called the Thundercats. And it's a great lead into this next question, which is if an organization is going to bring somebody in to lead AI transformation, how do you make sure that that person's not just a mascot?

Speaker 2 (12:52):

How

Speaker 1 (12:53):

Do you empower them to actually make transformation happen?

Speaker 2 (12:57):

Right. And that's a hard one. And we still try to make sure we evaluate and define what that looks like every day. But you're right. It is one thing to bring in a cheerleader or an evangelist. Some organizations might actually find those useful. And maybe that's exactly what they need is someone to go in there and just cheerlead the adoption or the transformation, but provide no real meaningful change. That's up to the org. Our Thundercats were already helping in that effort. They already had the mascot and they were doing. So the organization knew they didn't need to have a mascot come in. They didn't need to have someone be the face of this because our CFO's effort was already the face of it. So when you're bringing in that external resources, you want to make sure that you're able to extract from them or put the onus on them to identify those KPIs and those ROIs.

What does AI transformation and adoption look like here? How can we track it? How can we make sure it is meaningful so we can hold all our resources accountable? And that's really, again, up to the organization, up to the leadership, but it's definitely something that it needs to be talked about upfront. It is not simply coming in and having someone named an advisor who just points and says, "Do this and that and thank you." You don't necessarily want that, and we didn't want that here either. The

Speaker 1 (14:20):

Keyword there was accountable. Someone who's accountable for actually demonstrating this is working.

Speaker 2 (14:25):

Correct.

Speaker 1 (14:25):

What do you need to empower this person with in order to make that happen?

Speaker 2 (14:30):

Yeah, a clear mandate. What is it we're trying to achieve? Here at every level we have our mission, we have our strategy, we have our tactics on how we execute those things. Being able to articulate what you're doing at every point in that is really where we can identify and hold accountable. And that person as well needs to hold accountability to what their mandate is. I would say in our case, you have your six month, your one year, how many use cases do we want to bring in? Do we have an ROI target? Hard blue money, green money kind of targets that we want to enshrine in some kind of mission statement. These are all valid ways of kind of hitting that mark. And it really depends on the organization, what you're looking to achieve. Honestly, our AI transformation can and will look significantly different from other AI transformations and innovation.

It's really going to come down to the organization and what it looks like itself.

Speaker 1 (15:35):

Let's break down the transformation word. This is just one of those fluffy umbrella terms, but you are being asked to be accountable for actual metrics and

Speaker 2 (15:46):

Basically

Speaker 1 (15:47):

Proof points of what transformation means. Yeah. Whenever

Speaker 2 (15:49):

We transformed. Right.

Speaker 1 (15:51):

So give me some example of what actually is transformation. What is something that you and the Thundercats have successfully said, yes, we can point to that and say we have achieved AI transformation.

Speaker 2 (16:02):

And that's a really good question because transformation is not simply, oh, ROI. That doesn't mean you've transformed anything. And it's also not adoption. And we ourselves and me have been conflating them maybe this whole conversation. AI

Speaker 1 (16:19):

Adoption versus AI transformation.

Speaker 2 (16:21):

Two different things. And even capturing KPIs versus ROI. That's not transformation, right? So when have we transformed? And transformation hasn't been introduced to the finance realm with AI. Finance transformation has been a thing for a decade now. It's something we've all been trying to do. So what does it look like? Especially when we're talking about AI. Transformation means that the work you were doing is no longer being done in any significant way like it was being done before. So that's the transformation is we have brought AI to bear against things and it has meaningfully changed the way we organize ourselves, the way we report to ourselves, and the way we track finance movements. So here at Insight, we would say meaningful transformation means that we're taking our finance organization from a process-bound movement that is transactional, moving things across the board and bringing to bear AI, automation, enterprise solutions to enable our team members to be analysis-based.

So instead of them just simply turning things over, they're above it observing it and they're turning the knobs. And that's going to be true transformation is when our team members here, our finance professionals are doing things that are totally different than how they do them before. Give

Speaker 1 (17:54):

Me one example, one concrete example.

Speaker 2 (17:56):

Yeah. Great. So AP, which I love. I have a soft spot for AP. I sit right next to our senior manager of accounts payable who's awesome. So I'm always talking and scheming, including with him on ways AI can transform its process. Sounds dangerous. It's great. And so one way we do that is until, wow, a couple months ago, invoices would come in and a team member would look at an invoice and they would code them and put them in the system. And it was a very high touch manual process. There of course was some automation involved, EDI, them electronically coming in. But at the end of the day, it was still very much these accounts payable professionals that were looking at invoices and looking at the payments, going in and out and making sure it got done. And what we've been able to do with AI is allow AI to do those analyses, to code the invoices, to understand what an invoice is and what it relates to.

And now the accounts payable professionals are just watching the flow of this and just understanding what is going on and what needs to be changed, what exceptions need to be accounted for, and how they can make sure that the AI has upskilled itself to account for those new exceptions. And that is something we are doing here today and has been part of our AI transformation journey.

Speaker 1 (19:18):

So an AP person came to you and said, "Hey, we're in this very manual process. It's tedious. No one enjoys doing this. You guys built together an AI solution that now it's essentially like having an intern." I hate to use that expression, but an entry level type position that's doing the majority of the work and then they're kind of just spot checking it and guiding it.

Speaker 2 (19:40):

I would love to take credit for this one, but it was already in flight when I got here. That's

Speaker 1 (19:44):

Okay.

Speaker 2 (19:45):

But I did tell our senior manager, if anything, I'm going to take credit for it if all else fails. But since it was implemented - On the record. Yeah. But no, this was already in flight because again, we were already building this muscle with the Thundercats. They were identifying use cases. And again, our AI adoption journey has not simply been find these singular resources and make the responsibility for AI on them. Everyone in finance at Insight is responsible for our AI transformation, for becoming that AI first finance organization. And so this was the senior manager of accounts payable, the vice president he reports to actually doing this analysis and understanding, oh, if we go and find ways that we can solve for this, if we bring in a system, if we create some kind of solution over here, we can meaningfully change the way we do AP here.

And so that was part of them doing it. And I think also a testament to our AI adoption success is the business itself kind of defining what that AI solution looks like.

Speaker 1 (21:00):

I love that you're serving the finance function. And I think it's so fascinating that this was a mission from the CFO. Right now, a lot of organizations are having to reevaluate their AI budgets because of the now usage-based billing on tokens. We've really changed the metrics behind AI adoption and really thinking critically about how do we use this the most effectively, efficiently to make use of our budget. So I'm curious from the CFO's perspective, maybe you can give us a glimpse into his mindset. Why was it important to him to be this AI led organization? Especially finance being you're a cost center, you're the people that sign off on whether or not where budget goes. You're also, not you, but the CFO is probably the hardest critic, I think, for most organizations to say AI is proving value. So you have some unique insight into what is a good AI ROI story for

Speaker 2 (22:00):

A

Speaker 1 (22:01):

CFO.

Speaker 2 (22:01):

Sure. And I'm not going to talk for our CFO. We'll get him

Speaker 1 (22:06):

On at some point.

Speaker 2 (22:07):

I see into his mind, but darkly. And he absolutely would be a great guest having had him on our podcast already. But I think you're asking what's the motivation here and the confidence behind going all in on this strategy, especially given the volatility and the ever-changing nature of this new technology. Where you're right, in three months it changes. In one week it changes. And betting big on that or setting down a mission of being the first in AI when it is constantly shifting under our feet, what was he looking at that gave him that confidence that said this is it? And I think he correctly evaluated and I think he correctly appreciates the truly transformative nature of AI on our work. But that does come with some insecurities that all organizations need to be comfortable with if they're going to go down this path. Like

Speaker 1 (23:07):

What?

Speaker 2 (23:07):

Like the ever-changing nature. So at the beginning of the year, the Thundercats AI operation team, we sat down and we said, okay, what are we going to focus on this year? What's 2026 going to be about as far as use cases, adoption, upskilling? How are we going to achieve that in finance so we can get to the mission? And so we set out these great pillars and goals and we identified the business technologies that have AI and that the business should be using. And then Copilot cowork comes and it changes our everything about what we expect the business team member to be able to do with AI. All of a sudden that has changed. What we expected ourselves and our IT partners to do. Now we have something that looks very agentic that's able to do it. And so the business teams can self-service. And that just changed our whole orientation.

And in three more months that could change again. So it's keeping your eye on the prize saying AI will provide this transformation. It will change the way we work, the way we organize around work, the way we think about finance, but we have to be comfortable and we have to be flexible enough to move with the technology as the value it provides changes where and how. And that's really been key. And that's something that our CFO and our senior leadership team has been really patient and understanding with us as all of a sudden we wake up and James, our CFO is like, "Have you seen this?" And we're like, "Oh no, why do you see this?" And us having to say, "Okay, things are going to change." And they have tolerance for that because they still do believe that the underlying technology and this underlying direction is correct and will be successful.

I've

Speaker 1 (24:51):

Seen that happen so many times where either myself or teams are trying to solve a problem with AI. And then a week later, Copilot, for example, rolls out with a solution that does the very thing we were trying to do. So

Speaker 2 (25:04):

It's

Speaker 1 (25:04):

Like, just wait.

Speaker 2 (25:05):

And there's a well-known concept in machine learning and compute in general called the bitter lesson, which is you will spend so much time trying to come up with a cunning or a clever solution. Writing your code this way, deploying your app this way, doing something to solve a problem. And then all of a sudden it can do it. You didn't need to do all that. And it is something that can lull you into paralysis because you're like, "Oh, hey, why should I work on this? Why should I spend this when it's just going to maybe solve for this in three months?" Maybe. Maybe. And the dictate from James, our CFO, is we don't wait. We go and try to capture it now. We try to move forward while we can, regardless of what we think might be coming down the pipeline within reason. And that's also, I think that clarity of action is really useful because it means no one can sit there and say, "Well, let's wait till tomorrow."

Speaker 1 (26:09):

That's a great point. And I was also just thinking regardless of whether or not the solution comes, you're learning something in the process. We're building the

Speaker 2 (26:16):

Muscle. So it's not being wasted. That's correct. And that's something. So again, we're telling people, "Hey, use Cowork as much as possible." We're a frontier organization, which means for a couple months there, we weren't subjected to any consumption billing on co-work, which means it was free. And so we were just telling the business team members, "Just use it. If you have a thought in the world, just use it." And so we were trying to build the muscle. And then all of a sudden it's released and us, like every other organization is subjected to that consumption-based billing. And all of a sudden we have a lot of very disappointed, if not angry team members who are like, "You told me to use this and now I have to worry about being efficient with it." And I think that is more, let's build the muscle while it's here while we can.

And then when that new tomorrow comes, that unknown pops up and we have to shift or understand it differently, that building the muscle was not wasted. The way we understand this was not wasted. It enables us to go forward with confidence and maybe shift a little in the process.

Speaker 1 (27:25):

Great advice. Cannot be understated. I want to ask you more about the tactical and practical ways to make this work. You have a team in-house that you're operationalizing around figuring out where can we re-imagine our processes and bring AI into those processes and create new workflows. I remember a time when we were throwing everything at Teams to try to get them to come up with AI use cases. And it was like no one could really imagine what this meant and how it worked. And then it was almost overnight, the use cases and opportunities were just drowning us. There are so many places and use cases for it. How do you go about deliberately finding the

Speaker 2 (28:13):

Right

Speaker 1 (28:14):

Use cases and making sure that these don't just become like a hundred different pilot projects that don't actually mean

Speaker 2 (28:21):

Anything

Speaker 1 (28:21):

At the end of the day? Or

Speaker 2 (28:23):

Just way too big and broad to ever meaningfully address or understand what you did at all.

There's got to be some kind of level. And that's truly where the Thundercats shine. And we're frankly crushing it before I got here. And that is with our upskilling and our education program. In 2026, we took that to the next level. So the AI operations team, the way we tackle stuff is we have three pillars that enable, energize and evolve. And in each one of those pillars, we have different functions that we execute and those are our tactics. So in energize, that's our education. That's where we do labs with our team members in finance. We show them how to use something. We do office hours. We make ourselves available for them to come and talk to us about anything. Showcases, people having these, "I did this with AI opportunities," and us helping identify that in any number of ways that we can provide our team members the knowledge and education to use AI.

And in the course of that, we identify use cases. They just bubble up. And it's a really great way to organically bubble up those use cases is by kind of giving the business that knowledge so that they can go and actually figure out what is a good AI use case. The enable pillar is exactly what it sounds. That's where we make sure that the business teams have all of the tools to self-service, to meet their use cases, governance and policy on how finance Can use AI in a responsible way and what that looks like and how to measure the outcomes. And then finally we have our evolve and that is the much more classic, let's go help build something with the business. Those might be use cases where we know it's a good one and the business just doesn't have all the technical capabilities on the tools to be able to build that workflow, that prompt, that skill.

And that's where the Thundercats will come in and use their 2.0 capability to solve for those things. And that's also where we work with our AI COE and our IT partners to make sure that we elevate those use cases that are not something for the business to do. There's still definitely a place for IT to be in on this and help service. And those are proper projects, but the use case would have never been discovered were it not for the Thundercats facilitating that. And that's very much kind of the three tiers at which we're able to identify use cases is making sure the business can identify them, self-service them. The Thundercats engaging with the business so that they themselves can find the ones that these finance teams might miss. And then lastly, engaging with our IT partners and saying, "Hey, what's the possible? At what point is this a category two or three?"

Speaker 1 (31:24):

Meaning how complex it is? How

Speaker 2 (31:26):

Much do we need

Speaker 1 (31:26):

To get IT involved?

Speaker 2 (31:27):

Exactly. There's definitely a. We're not shadow IT and we don't want to build one. There is a spot and a correct one for everyone at every level of this. And the Thundercats are very much helping identify and make sure those things are where they should be.

Speaker 1 (31:43):

What's been the most surprising discovery when you're asking for use cases or people are coming to you with problems to help them solve?

Speaker 2 (31:51):

Surprising is some. It's not surprising, but it's very interesting. It's the novel ways that our team members are able to use technologies, which I myself think I'm pretty good at. And we will go in and someone will say, "Hey, I was trying to solve this problem here." And I think I did it. And we'll go and see and we'll be like, "Oh wow, this is not only correct and awesome, but it's also novel in how they understood it and how they were getting something to work." Case in point, we do pricing analysis on all of the things we sell. Great pricing team that does that. And they have all these dashboards they look at that kind of tell them this is how good our pricing is. And we were told that one of the team members on that team had used one of our AI tools to basically have AI do the analysis.

And he was sharing that with the pricing team and they were saving hours of time because this AI tool that he kind of built was surfacing those insights. And we said, "Oh, cool. That's awesome. Let's go and make sure that we can 2.0 that. Let's go make sure it's as good as it can be." And we went in and we spent an hour talking with Jake and we realized that he nailed it. There was nothing for us to do. It was incredible. We were literally like, "Well, have you tried this? Or what about this?" And we're like, "No. Okay. Yeah, that is correct." Very impressive. And so that's maybe at this point, not as surprising anymore, just given how good our teammates are. I was going to

Speaker 1 (33:23):

Say, it's a good testament to how much training exposure you've given the team that they -

Speaker 2 (33:27):

That's correct. Where we're moving from this upskilling or this education to adoption. And they're adopting and they're adopting in novel ways with the tools that they have that are more than capable.

Speaker 1 (33:40):

How many problems,

Speaker 2 (33:42):

Quotation

Speaker 1 (33:42):

Marks, come to you, that are actually solvable with AI?

Speaker 2 (33:47):

Yeah, that's a good one. I'd say, and this is something we talked about early on, this is something I talked about early on with our CFO when we were establishing what success looked like here for an AI innovation advisor, is early on we said nine out of 10 of these use cases are not going to be AI.

Speaker 1 (34:03):

Why? We

Speaker 2 (34:03):

Have to be comfortable with that. Nine out of 10 of these. Well, why are we seeing nine use cases that are not AIs? Because when we go out to the business and say, bring us your AI use cases, the business, not unfairly, here's just bring me your problems. Which is fair because they're trying to alleviate some of these processes or ways of doing things that are just hard. And they're trying to say, "Hey, can AI solve for this?" But nine out of 10 times what we're seeing is not an AI opportunity, but rather an automation opportunity, a broken business process, or a change management opportunity, or an enterprise solution that might already exist and the team was just unaware of it. And early on, our senior leadership team said, "That's fine. We can do those too." And so the Thundercats do take those on, but nine out of 10 of them are not going to be AI.

Now, once you do go and take those automation opportunities or those broken business processes and solve for them, that's where you find your AI opportunity. Once these problems, these issues, these opportunities are surfaced and solved for in a non-AI way where you can't solve for them any other way, and so the solution is correct, that's when you say, okay, now I can use AI on this. And so that is just more motivation to do these things because that will get us to the AI opportunity.

Speaker 1 (35:35):

And

Speaker 2 (35:35):

That's why it's in scope.

Speaker 1 (35:36):

Yeah. Well, it's even a win to fix something that wasn't working anyway, whether

Speaker 2 (35:40):

It's it or not. And that is another big thing that the Thundercats take very seriously is we don't say no. It doesn't mean we do everything, but we are very customer focused. And in that we just take on the work. And that may be nine out of 10 instances where we don't do work. We just go and make sure that it's routed correctly. And we call that routing. It's in our evolve pillar where if someone comes to us and we say, "Oh, you know what? Actually, that's probably a good business intelligence dashboard opportunity. Let's go have that conversation with that team. Let's make sure that those requirements are translated correctly and that you're getting the help that you need." A lot of times the business teams are just kind of out there in the wilderness looking for the relief. And if the Thundercats can just help them get to where it already exists, that's a win.

Speaker 1 (36:30):

I see that's so helpful for organizations that are long established. They've grown maybe very large, very quickly. And you find out all these broken pieces. It's kind of like having an old house. The floorboards are creaking. There's flood over in that room. And it's like you have a team who maybe they can't solve your problems, but they can point you to the people who can, the expert in plumbing, the expert in flooring, et cetera. So let's get to the actual process for when you do agree on a use case, you're going to work. Let's say I come to you, I've got an example of a workflow or something that needs to be remit. What is that process? How do you now work with me to make that happen?

Speaker 2 (37:07):

Sure. Number one, again, from a customer obsessed perspective, we're here to get the outcomes that you need for your team to execute and do its job or transform and move into a different way of working. So we're really, really cognizant and conscious to say, okay, we're not going to come in here and just wipe all the things off the board and say this is what you're going to do and be over prescriptive. We like to work with the teams to make sure that we're getting the outcomes that are useful to them. Not necessarily this dogmatic way of saying this is what needs to be done. We set up a cadence. So they come to us and number one, we have a pipeline where we can see all the work. We seek to make that incredibly transparent and communicate where we're at in that effort. And then depending on what the ask is, we try to make it very clear this is how we're going to execute on it.

Again, we have organic opportunities that come in from a team member level. And in that case, we go and help make it important to their leadership that this is a good use case and that they should work on it and help them articulate and scope it and say this is the technology we're going to use or they should be able to do this themselves. To our active assessments where we go and engage directly with the senior leadership on a team and say, where are your pain points? And let's see how we can evaluate those and right size them and then direct them in the correct place. Is this a self-service? Is this a routing? Or is this a category two or three where it's like, our AICOE can probably do this. Let's work with you on communicating to that. And then we continue to track that and make sure that everything is executed and that everything is communicated correctly.

Speaker 1 (38:56):

I'm going to ask you an unfair question. How many projects might you work on at a time and how much time will you allow to work on a project to control the scope?

Speaker 2 (39:04):

Yeah. Depends on the project really. I'm going to give you an unfair answer. I know you're going

Speaker 1 (39:10):

High

Speaker 2 (39:10):

Demand.

No, and that is something that was very important to us early on and something that also went into understanding what success looked like here, is making sure that we had a healthy pipeline. What projects do we have in our pipeline? What is our execution cadence? How can we prove out that there was some kind of use to doing that? And so anywhere, depends on what your execution cadence is, but we expect to have at any given time, 10 to 15 real useful use cases that we are tracking and trying to solve for at any given time. Honestly, we like to be fast, but it also depends on what the actual solution's going to be. If we're working with our AI COE partners in IT, we know that their project timelines are a little different than what we can execute on. When the Thundercats themselves go in and say, "We're going to help you solve for this," we really seek to have those done within a month.

Speaker 1 (40:10):

And explain real quick, I know you mentioned that Thundercats is the dedicated team within finance, but just for someone who's listening is like, well, he's mentioning Thundercats. Wh does this team actually look like? How many people are we talking about? And do they have jobs outside of this AI transformation? They're dedicated to AI. That's correct. So how big is this team?

Speaker 2 (40:28):

Who are they? So currently they're here. We have a member in Europe, and then we have another team member in Manila in the Philippines. We service our global organization. Our European team member does not just service EMEA use cases. And same with our Manila one. So we all work in tandem and there's six of us at this point and an intern, Josh, who is integral to the team and we love having. And they are full-time dedicated to basically executing on our pillars that energize, enable and evolve. This is what they do, this is what they think about. But coming from within the business, coming from internal audit, from accounts receivable, accounting, HR or AP really enables them to better and more quickly understand where the value is and what the valuable use cases are and also where all the bodies are buried. I can't tell you how useful it is when Tim, one of our Thundercats, we go and evaluate a use case and he's like, "I know this process.

I'll talk to you about it later. We'll understand what all," because there's a lot of context that goes on around it. And so it's really, that's where the Thundercats shine is being able to say, "I actually used to do this. Let's figure out there's many things that could go on here. And AI might be a very little part of this, but it doesn't mean we can't solve for everything else around it while we're at it."

Speaker 1 (42:00):

So these were teammates that existed at Insight. They had different roles. They were already in

Speaker 2 (42:04):

Role.

Speaker 1 (42:04):

Yeah. They have all that context. Now they're focused on AI versus what I think so many organizations face, which you have people that have their day job, their primary job, and are trying to figure out the AI on top of their job. Now that you've seen what the ThunderCats have accomplished and continue to do, do you think that this is a model that should be replicated across business

Speaker 2 (42:26):

Units? I think it's a great model. It really depends on your organization. It depends on what your finance - It

Speaker 1 (42:31):

Feels like a luxury.

Speaker 2 (42:32):

It is. And it's an expensive luxury too. And that I think is also a testament to, number one, how we can find value in what we're doing, but also a testament to the seriousness of the CFO's mission here that he set for us. If we are truly going to be the number one AI finance org in the Fortune 500, that's going to take a serious investment and a serious signal of how serious we are to that. So you're right. And that's why it comes down to it depends on the organizational tolerance for it. I would say if you truly believe that AI is this transformative moment, that this will truly change the way we do business, whether it's finance, HR, legal, sales, I'd say it's going to prove out your investment in it. It's going to prove out that uniqueness and that luxury that you might have to afford in a specialized team.

But that's not to say that the champion model can't work. And we do have other parts of our organization that do use the champion model, which is someone that is still in role, but also is the AI first person, the one that is maybe attending a Thundercats lab to upskill and become more fluent so they can translate that back to the organization. And we have teams where that has been very successful. So it really depends on your size, your situation, and where you want your organization to go.

Speaker 1 (44:00):

I'm going to challenge you with something because I think this is an important time to have very honest conversations about the value of AI. The CFO has this sort of mission to Mars, right? Sure.

Speaker 2 (44:13):

We're

Speaker 1 (44:13):

Going to become AI first. That's great. People can rally behind that. But how do you translate that to, I just want to work at a great company and do my job and feel good about the work that I'm doing. What have you observed to be the most meaningful elements of this sort of mission to Mars mentality?

Speaker 2 (44:33):

Yeah. It's making sure that our team members know that there is more for them to do or there's more meaning in this AI transformation, in how they work. And that's super important, is not saying, "Well, I just think this is a RPA 2.0 tool that will help me get leaner and more efficient. And it's not really about AI or I truly think that AI is going to transform the way we do everything and the way we think and talk and eat and all that. And therefore everyone else should feel dustly." No, it's more about going to your team members and saying, "Come what may. There is going to be meaningful work for you to do here." And if anybody's worked in finance, in a cost center, there's so much work that we can't get around to doing with what we have. And so it's making sure that your team members know, "Hey, we're going to try to solve for this thing with AI.

And I know it's a thing you do, but that's because there's something that we need you to do that isn't this." And you really have to. And our CFOs and senior leadership team is really good at kind of communicating that out and helping even our team members at every level understand, well, what does it look like when we get there? What does it look like for us and why?

Speaker 1 (46:03):

I love that. I'm going to close with you for a rapid fire round.

Speaker 2 (46:05):

Here we go. I

Speaker 1 (46:06):

Know.

Speaker 2 (46:07):

This

Speaker 1 (46:07):

Is the pressure's on. Give me a good sound bite here.

Speaker 2 (46:09):

All right.

Speaker 1 (46:10):

Worst advice you can give somebody in terms of AI.

Speaker 2 (46:13):

Ooh, worst advice. We see this all the time and we saw it not so much anymore, but when AI came out and there was that ChatGPT moment, and we saw companies really trying to take a hold of it. And we saw a lot of pressure from the top, not necessarily here at Insight, but we see it all the time that companies we as Insight go and talk to is there's this pressure, use AI, use AI. And it's just this really blunt pressure. And so we will see team members that are trying to solve for something. And we hear a manager say, "Well, have you tried to use AI?" And maybe the sentiment is correct, but the advice is not helpful at all. And one test we try to give our teammates to apply to this is swap out AI for the term the internet. And if it still sounds absurd, it is.

Have you tried the internet? So imagine, yeah, if you're sitting there and you're like, "Hey, I'm trying to do this or that in my daily workflow." And someone says, "Have you tried using AI?" Just take that out and say, "What if they said, have you tried using the internet?" And if it still sounds absurd, it is.That is the test. And so we find that, again, we're seeing it less and less as every member of the organization understands what AI is and is not. But that is some of the worst advice is just to say, "Well, have you tried AI?" It's painful. Yeah. Alex,

Speaker 1 (47:42):

I don't think you understand speed round, but I'll give you a pass because that was a great answer.

Speaker 2 (47:45):

No, no, you go past. I guess so.

Speaker 1 (47:48):

All right. Similar to that, what do you believe is the most overrated or unhelpful AI metric?

Speaker 2 (47:55):

Oh, geez. There's so many of these. I think one thing we see is usage. Just raw unadulterated usage without actually understanding what value is therein. That was quick. That was it.

Speaker 1 (48:12):

If you were to start from zero, what's the first workflow you would go fix?

Speaker 2 (48:17):

That's a really good one. In my own or in the organization itself? You can be

Speaker 1 (48:25):

Both.

Speaker 2 (48:26):

Really? I think I've really nailed it at every level so far, so I'm not sure I would really go back and do anything else. That's only because I can't think of one.

Speaker 1 (48:39):

All right. One thing a leader should do before buying an AI tool.

Speaker 2 (48:43):

Oh, yeah. Understand what you're trying to do with it. We see this all the time with our customers here at Insight where there's such pressure to use AI. And so they're just going to bring this thing in and it's going to AI. It's like, well, what'd that look like? What's that supposed to be? So it is really going into making sure you understand what AI is and what it is you want it to do in your business. That might not be transformation. That might be just a specific use case. And it's fair, but you need to understand and right size that expectation.

Speaker 1 (49:16):

When it comes to AI transformation and you're deciding between a dedicated owner or distributed ownership, which wins? Oh,

Speaker 2 (49:23):

Again, comes down to the organization, but there's also a why not both. Depends on your tolerance. I'm a big fan of distributed responsibility and ownership with a centralized governing guidance. So build the walls high and let people work within them. So I really like the balance of both of those. You really do not want to. We have so many use cases, especially as a cost center where we have use cases that we may work on and HR and legal have those too. And if we were so distributed, if everyone kind of owned their own piece of it, we would never find the ability to lump them all together and solve for them completely.

Speaker 1 (50:07):

Last question. How do you know when you're truly in AI transformation mode versus just doing pilots?

Speaker 2 (50:13):

Ooh. Yeah. When people stop talking about AI.

Speaker 1 (50:17):

Truly.

Speaker 2 (50:18):

Truly. That's when my job goes away, right? When we've truly transformed either a team or a team member, they are no longer doing AI, they're just doing their job. They no longer need an AI innovation advisor. They no longer need a Thundercats because they're all doing the AI. It's just another thing. And that is truly the metric of transformation and success with AI is where it is just part of how you do business. Well

Speaker 1 (50:45):

Said. Thank you. Alex,

Speaker 2 (50:46):

Thank

Speaker 3 (50:47):

You so much for your

Speaker 1 (50:47):

Time today.

Speaker 3 (50:48):

It's

Speaker 2 (50:48):

My pleasure. Thank you.

Speaker 3 (50:50):

Thanks for listening to this episode of Insight On. If today's conversation sparked an idea or raised a challenge you're facing, head to insight.com. You'll find the resources, case studies, and real world solutions to help you lead with clarity. If you found this episode to be helpful, be sure to follow Insight On, leave a review and share it with a colleague. It's how we grow the conversation and help more leaders make better tech decisions. Discover more at insight.com. The views and opinions expressed in this podcast are of those of the host and the guests and do not necessarily reflect on the official policy or position of Insight or its affiliates. This content is for informational purposes only. It should not be considered as professional or legal advice. I