Alright. Looks like we are live. Russell, Steven, you hear me okay? Yes.
I can hear you.
Alright. Well, welcome. I am genuinely very excited to be here today. My name is Michael Shields.
I'm the head of procurement and strategy at Tropic. I'll introduce our esteemed panelist shortly, but suffice it to say they're both on the finance end. And I think you both know this about me, but I even though I've done procurement my my whole career, fifteen years at this point, I've often reported the CFO. So I like to think of myself as a finance guy in some ways.
I don't know if you accept me or not in that sense, but that that certainly is the truth.
So we have we have forty five minutes today, and the goal is that people listening in, they're gonna come up with some ideas on how they can harness AI specifically, you know, for finance and procurement teams. But, you know, to make to make their job easier to, you know, the process is smarter, the results the results more impactful.
And, you know, I wanna share a lot of examples. So, you know, Russell and and Steven, hopefully, you came prepared for that. But I also recognize that, people are at varying levels of acceptance when it comes to AI. I actually heard someone compare AI to, you know, the stages of grief. Right? Some are in denial.
Some have accepted it. Others are embracing it. I, for one, am excited about it. Absolutely.
There's there's some trepidation, it but it's blowing me away every time I see it. But I also recognize I'm not an expert yet. So if anyone listening in is is kind of on the fence, I I would say a good goal would also be to start experimenting more with it, you know, testing it out, trying it out, etcetera. So, Steven or or Russell, any goals, you know, from your perspective that you would add to that?
Sure. Yeah. From from my end, I think on the, stages of grief thing, I I like that a lot.
I personally have been back and forth between stages depending on what the AI solution is. Some make me more nervous to use than others. But, overall, I think we're starting to get some traction in, like, actual usable solutions for what we do. And, generally, I am excited about, kind of the, the next frontier for us.
Awesome.
Yep. I agree. I'm very pro AI, but also in that category of, like, what do we do with this emerging thing called AI? What even is AI? How does it fit in finance? How do you balance risk that it might represent an opportunity? And so I just think there's all kinds of interesting stuff we can get into today about this.
Okay. Awesome. Well, it'll be an exciting conversation either way. Steven, let's start with you for introductions.
You've worked with in finance at a handful of companies. You're now the FP and A manager at CompanyCam.
What is something that this audience should know about you above and beyond that?
Man. Well, first and foremost, I am a lover of dogs.
Before my career and after my career. That is, true. But something more related to me and work is I really enjoy building things. I love building processes, and I love, helping to, like, bring disparate things together to more of a cohesive kind of thing. And so I really enjoy just sort of playing around with the building blocks that that we get to use and, create new stuff for the company to use.
Awesome. Love it. Well, we're glad you're here. Thanks again. Russell Lester, CFO extraordinaire. That's, didn't know if you know this or not, Russell. That's the title when you get when you've done the CFO thing at least four times.
So I'm sure your accolades, your accomplishments are many.
I know some of those. We know you're a diehard Star Wars fan.
Here's here's your first question. If the audience could ask you one question about yourself, what question would you recommend they ask? And more importantly, how would you answer that question?
Maybe they would ask, why am I such a die hard Star Wars fan? Or why is this poster hanging behind me? And it's actually because of my family. So my son gave me this as an early Christmas gift that he was so excited about.
He couldn't wait, and he came downstairs. He'd set up all the lighting. And so a gift given from the heart that someone is so excited to give that they can't even wait till the day of, that's a pretty awesome gift. And so it encapsulates, both the love of my family and love of all things Star Wars.
K. And what's higher? Love of family or love of finance?
I'm sorry.
Family first always.
Ah, okay. Well well, look, I'm here.
Finance and procurement, we're we're working together a lot these days.
The partnership is really, really strong. I would say it's definitely strategic. I I know in the world of tech, there's a lot of finance leaders who are thinking about procurement, maybe not for the first time, but certainly from a more strategic point of view than ever before. So that's why, you know, I'm moderating today. I'm gonna chime in here there with some ideas as well.
But but let's let's dig in. Cool?
Yes.
Steven, let's start with you.
Just to kinda kick the, you you know, get things going here. I'm hearing a lot from, you know, finance leaders who I meet with, procurement leaders, that their teams, their resources, they're limited right now. The margins are, you know, the demands and the expectations are really, really high.
The demand, you know, the need for margins to be high has ever been. Maybe talk to me like how does that hold true for your company, and just kinda give a lay of the landscape. Let's set the stage there.
Yeah. Absolutely. So CompanyCam is, I think, going on about ten years of existing at this point. We're still in a pretty early stage of the business experiencing a lot of growth.
And, so one of the things that's really important is growing at the correct pace, and that's also, not just, like, our revenue or top line numbers, but that's headcount and expenses and things of that nature.
Just because there is money there doesn't mean we should spend it. We always wanna make sure that we're getting the most bang for our buck. So we try to run as lean as we possibly can having, I think by the end of next year, we'll have a finance team of, like, six people.
And for quite a while, it was just two or three. And so getting as much as we possibly can out of the people that we can is extremely important. And I don't mean working people to the bone. I mean, allowing them to do things more efficiently and be able to accomplish the same quality of, what they normally produce, but taking less time.
Yeah. Very interesting. Russell, anything you would add to that?
Yeah. I mean, Steve mentioned people, and I also think about the processes and the tools. Are we getting out of the processes and tools that we're investing in what we hope? Like, what are we hiring these different things to do?
Because every team, ours is no different. We are faced with that pressure, the capital pressure, the runway pressure, the time pressure.
Office of CFO is taking on new duties every single day with fewer people. So how do we get it all done? How do we get it all done well? We need scale, and we need efficiency. So I think I wake up every day thinking about those two things. How do I balance those?
Yeah. That's really interesting. So what I hear you guys kinda saying is not that you're necessarily putting more on your team's plate, but you're trying to find ways to be more efficient with the the smaller size or or the limited resources that you do have. Right? But, so so, Russell, let's let's double click on that a little bit and maybe tie it into AI, which is obviously our topic today.
Mhmm.
You know, as we mentioned, you're a four time, you know, CFO. You had a few jobs before, you know, that. Obviously, you you've been around the block once or twice.
Maybe, you know, looking back, how does this AI concept compare to other exciting developments or progressions that you've kinda seen, in your career? Like, may like, how how does it rank? Like, you know, I don't know. Just yeah. Double double click on that a little bit.
I mean, you think all the way back to the days of, Microsoft Clippy, which some people probably don't even know what it is. And it was this little agent trying to predict what you might need to do, but it was really just a choreographed and rehearsed, clip art that was basically intervening, sometimes not very helpful.
You fast forward to things like chatbots.
Right? And you think about the way people had to program in those chatbots, and they were only as smart as the programming was. And now today, you have very interactive AI capabilities that sometimes make you even wonder, am I talking to a real human or not? Because it's it's so real.
And that's very top of mind. But what about for the office of CFO? What about for financial tools? I think financial tools have been much slower to move.
The back office is often not the first place to be invested in. And so I think it's evolving and emerging, but it's leading with, data first. Data first solutions are leveraging AI more so than maybe process flows. But I think we're we're reaching the point where this spectrum of, like, identify, classify, recommend, predict, act, nudge.
We're seeing solutions crop up that are intervening across that continuum.
Yeah. And and, Steven, maybe let have you weigh in here as well. But as you think about this sort of impact that that it certainly can have because it it it is incredible, maybe let's dig in on some of those pain points, you know, that you're currently experiencing or that are taking a lot of time because that's arguably, especially in in the world whereas, you know, you both started talking about in the beginning where, you know, resources are limited, that that's that's probably gonna be, you know, the the easy use cases or at least the first use cases.
Yeah. Absolutely. I think, we in finance have so many cyclical activities that we have to do at certain cadences and, under certain timelines that having something set up to be able to automate a lot of that is a massive help. Something that I think would be, worth exploring for us in the new year is, the, AI functionality in our FP and A software is going to be able to help us with, variance analysis, root cause analysis, and things like that that really just take time out of your day, clicking through, getting to the transaction list, looking at things.
That's valuable information, but it would be so much more valuable if we can spend more time acting on the insights rather than actually looking for the, you know, several little transactions that flow through here or went somewhere else. Just saving the time there, I think, is is probably the most impactful to me and my team.
Yeah. So you you talked about various analysis and and, you know, anomaly detection, root cause analysis.
Do you have, like, a tangible example that that maybe might might be worth exploring?
Yeah. I think that, so we have a solution called Pigment, and that is our FP and A software. It's it's fairly new with their AI functionality, but something that's going to be really helpful. We just have to set up all of our, data to give it the guardrails with which to to act in. But prior to that, it's really just setting up, sort of checks and flags to have the system automatically ping us, and that happens pretty much every single month. I think just this past month, there were a few transactions that, were automatically flagged for me that I was able to review that I may not have seen until significantly later in the day, and I was able to have a conversation with the, people that had the context around that so that I could understand what happened, on that team and be able to provide my commentary and, explain what what was transpiring during the month of close.
Interesting. I you know? And I start to think about, like, how this might tie into the world of procurement some more, and and I know we're probably, for a lot of companies, wrapping up, you know, budget season. So the example you shared almost felt a little bit reactive, which which is good because you wanna do that root cause analysis to kinda understand, and I think that's really important.
But but also from a procurement point of view, I can see, especially in the world of f p and a where you sit in, this concept of of spend forecasting, right, of not only kind of, you know, understand root cause analysis, but, you know, almost helping so that those issues don't pop up in the future. Because I I think about in an old world where you're forecast not your budgets, you know, you need that data and insights to know, okay. Do we forecast an increase here? Do we forecast a decrease here?
Right? And and how, you know, I see a huge opportunity where I AI tools can help predict cost, avoid surprises, that sort of thing. Is that does that resonate?
Yeah. Absolutely. I like to think of it as a sanity check.
Having the initial forecast that comes out of the automated solution is great to have to check against the assumptions that we have, to check against the, forecast that we've produced. Because right now, truthfully, it's just not quite at the level where I completely blindly trust that it's going to, create a significantly better forecast than, what we've put together. But maybe it will catch something that we did not see in the underlying data. So I think that's that's kind of where we're at. It's it's hopeful for the future. It's great for, gut checks, but not exactly something that we're prepared to completely rely on as of yet.
I I hear that a lot. Right? Right? Where it's like, hey. AI can take you some of the way, but then there's there's a little bit of quality assurance or maybe just a lack of trust, you know, for for you know, at this point. I think that'll improve over time.
Russell, may maybe, you know, going back to you a little bit and and, you know, you you talked about some of the trends in finance.
You know, maybe there's a use case that you could share. I don't know if it's regard to forecasting as, you know, Steve and I were talking about a spend analysis. Let's go to you for a second.
Yeah. I mean so remember in finance, we're focused on the revenue side and the cost side. On the revenue side, a lot of those tools are more top of mind emerging. Clari, for example, very well known for their AI capabilities to predict where will the the funnel, where will the pipeline land, what is the likely trajectory to the day.
And you can click and drill down, and you can ask questions of it. And so in the continuum that I that I talk about where you're identifying, you are classifying, you're recommending, it's kind of flying in that territory, even in the prediction. It's not really acting or nudging. Like, you're not actually changing workflows in a tool like that, but it's surfacing those insights so that you can better predict revenue well.
Similarly, such as in in Tropic, of course, I use Tropic every day, you can ask questions of Tropic about contract data. Well, there are tools that let you ask questions of your general ledger data in a similar way.
With FloQast connected to NetSuite, for example, two tools that people on the line are probably very familiar with, they are beginning to interact with each other in ways that you can actually, interrogate or ask questions of your general ledger. And it's all in service to trying to more quickly identify your variances.
You know, what are the key, suppliers that are causing variances month over month, the key expense categories, what do what do the trends show, what does the contract data indicate, is likely to occur. So and even things as simple as Excel GPT, which apparently is a thing. I haven't myself used it, but my team has used it. Yeah.
And it's it's just amazing how quickly this stuff is accelerating. And so the best thing we can do is just, like, begin to try this stuff out, interact with it, learn it. The you're not going to just, wake up one day knowing how to use it all. We have to try it.
We have to lean into it.
Well, it's interesting because, you know, you're talking about, several brands out there that are well known, especially in the world of finance.
And I don't hear you saying that these core tools are gonna go away or that they're gonna be replaced by AI. But, you know, for your ERP system, your visualization tool, or whatever it is, it's almost like you're saying AI functionalities being incorporated into it. This idea of Excel GPT, that that does doesn't sound like you're taking an Excel file and putting it into, like, a chat GPT, but it's it's embedded right in. And it's it seems like what you're suggesting I don't wanna put words in your mouth, but that's kinda more the future rather than a tool stack being completely replaced.
That is correct. And I refer to it as AI in context.
Because AI for AI's sake or AI randomly floating out here Like, we all have chat GPT or some version of that pulled up, asking everything all day long. Is it the new Google? I don't know. But, that's different.
Right? AI in context means as you're living out your day as a financial practitioner in the workflows that you're doing, as you're trying to pay the bills, collect the receivables, balance the general ledger, close the books, perform variance analysis, you know, prepare summaries for the board, identify key levers and opportunities through all of those different activities that we all do each week, each day.
AI in context surfaces within those tools as a way to help you out and save you time. So the companies that are doing well are beginning to present those to the users as, more than it's gotta be more than novel. Right? It's actually gotta save time and do something beneficial.
Otherwise, it's just like Clippy.
Yeah. That resonates really well with me. Steven, I'll come to you in just a second. But an example that comes to mind is, yeah, I I I love, you know, logging into, you know and there there's a couple I have access to where it's, you know, a stand alone AI solution. Right? But if I if I go to this solution and, you know, let's say I wanted to benchmark a, a contract that I received, you know, and I said, hey. You know, chat GPT, am I getting a good price on and then I insert name of supplier and I'm buying two hundred licenses.
You know, it has the ability to go and and scan the web really quick, but a lot of that information is not available on the web. Right? It's it's it's proprietary kind of, you know, contractual information. And so it doesn't have the context to provide a relevant answer or recommendation.
However, though, is and you you know, you mentioned Tropic. If I log in to Tropic and they have their AI tool and I and it says, hey. You know, am I and I put that same question in here. All of a sudden, the data that it can pull from, you know, kind of behind that, you know, firewall, so whether it's Excel or or Clari or NetSuite or or Tropic, you know, then all of a sudden, the context is so relevant and so helpful.
And and that, in my mind, is a very, very powerful example of AI that is going to and already is changing how I operate.
Steven, you wanna pile on there at all? Absolutely.
I think I completely identify with that, and, using Tropics dot ai is a perfect example. So, I don't know, maybe a month or two ago before the Tropic dot ai system launched, I would occasionally try to use ChatGPT to get some insight into some of these vendors.
We get vendor, or we get requests to spend with vendors from all across our business, and it might be vendors that I've never heard of. It might be vendors that I'm unfamiliar with the pricing structure.
Things can just be kind of complex, and so trying to boil it down to what I actually need to know can be difficult through either ChatGPT or just googling around.
So just the other day, I was using the Tropic dot ai when we had a new, request for a data analytics vendor pop up. I was able to ask it those exact questions that you just said. I asked it, like, what the pricing model looks like. I asked it if we were getting a good deal. And most importantly, I think to me was, I asked it what competitors were out there in that space and what they did differently between the solutions to help us dig in a bit more and try to see, is this even the vendor that we actually wanna work with, or is it just the most well known one, that the team has has chosen?
So it's definitely saved me a lot of time googling, to get that actual contextualized answer.
Yeah.
And this is a huge area. This idea of sourcing or or understanding supplier, I think for in the world of procurement, this is going to be one of the most immediate and impactful concepts. And it'll, in fact, go to market teams. It'll I think it's gonna upend, honestly, you know, a lot of the market right now because, you know, how have we historically bought?
You know? You know? It's like, okay. Well, we're familiar with certain tools because we've used them at other solutions or maybe because they're top of mind from a a marketing perspective.
But when you can go say, hey. Who else competes here? And then you asked and then follow-up questions like, hey. Compare and contrast these two providers.
Right? And you're not necessarily relying on marketing material for that or, you know, just like, who else should I consider, you know, for companies like me and then disable to, like, all in what it knows about you? Like, that is incredibly impactful. And I would say that that's something that hasn't historically been done well because it's been so time consuming.
Absolutely amazing.
Obviously, there there are limitations.
Russell, you kinda talked about this a little bit where, you know, it's this broad spectrum. Right? It's identify, classify, recommend, predict. I can't remember everything.
You know, act, nudge. You know, Steven, you talked about, you know, how, you know, there's this quality assurance. Let's let's double click on not only what areas are most helpful for finance and procurement teams today, but but maybe where there's a little bit of of skepticism. And and and and let's, Russell, we'll start with you.
Yeah. Well, the skepticism comes from fear. Right? The fear in how is my data being used, and what type of intervention could happen behind the scenes that I'm not even ready for. Right? And this is why the continuum is important because I'm not sure financial practitioners would worry about AI being used to predict.
Because a prediction is right? Linear regression is a form of prediction.
And so we've we've used, you know, Monte Carlo simulation models in the finance world have been used for a long time. Is was that AI? We just didn't call it that. It's running a ton of scenarios simultaneously to come up with a prediction.
And so prediction is not the issue. I think once AI gets into the category of recommend, people are like, okay. Okay. Hang on.
Don't yet do the action that you're going to recommend, and that's the nudge or do. And so I think you see financial tools are accelerating around the recommend and predict.
Now the nudge or do could be coming where we actually are drafting recommended terms or we're actually drafting an initial variance analysis. Like, generative AI can take your general ledger data and actually draft variance analysis reporting for the board. But you still need to go in there and, like, put it in context of your business.
Right? So we'll never leave the need for human judgment. Like, we should not worry that AI is going to eliminate financial practitioner jobs.
What it's going to do is allow us to focus using our heart and our, our mind and our heart more to, do the things that we're actually find more enjoyable and we're hired to do, which is to help the company make decisions. Because AI can't make the decision for you. We have to use that judgment. It's just saving us time.
It's just speeding up the process of ingesting all of those data points.
Can I push you for an example here? Oh, and then go ahead, Steven, and then we'll come back. Go ahead.
Yeah. I was gonna say, I think, it kinda goes back to what you were saying earlier about learning to use the tools. We're not gonna wake up one day and it just works. So if we can learn to use the tools yeah. I also I agree that we're not at risk of making ourselves obsolete with these tools. I think that learning them is just the next skill that is gonna become eventually completely required in order to function at the level that is expected in our types of seats.
I am I'm not saying you're wrong. Obviously, I don't know. I because I agree with a lot of what you're saying. I I do think that there will be some entry level or more, you know, basic jobs within both the procurement and the finance world that can easily be automated.
Right? But that but that's not necessarily new. That's just you're taking something that requires a lot of kinda human intervention or or, you know, time consuming. You can quickly automate that.
But I I think that strategic roles, the FP and A, the, you know, the the the true accounting, I completely agree with you. But for either one of you, would love to you know, as we talk about the spectrum and the sweet spot of, you know, using AI to identify, make recommendations, nudge, you know, would love would love examples here if you have any.
Yeah. I mean, I have several. One would be you're rolling your headcount forward period over period. And headcount is a very tricky one in the world of finance because people come, people go.
AI can detect these employees left the organization. These were newly hired. You have these open roles, but this open role looks like the one that was just hired in the past that was so painful to navigate manually.
With AI, it can say, hey, It looks like this role is the same one as this one.
Should we just, remove this open role because it looks like it's been higher? That would be one example on the headcount side.
Another one that's actually pretty funny, I need to come back and mention it was, I was requesting an investment from you as my procurement partner at Tropic, and we went in Tropic. And you asked why on earth would I buy solution a b c?
And what did it I actually used those words.
I I I meant it to be, like, very much, like I'm not trying to be formal here, but I I remember this. We're on a we're on a call together, and I I use those exact words. Like, why on earth?
Anyway Why on earth?
And so through natural language processing, you know, and the magic of AI, it answered you by providing the business support that I had provided it as part of the part of the request, which was a very fun funny and fun moment on the call, not only to see the power of AI in context through Tropic, but you and I as parts of this approval ecosystem being able to quickly, you know, get get our hands around that and answer it, in context was just it was magic.
Yeah. I'll I'll come to you for Steven for, you know, for an example in just a moment, but I I completely agree. That was one of those moments. So I'm sure we've all had them where we're we're we're utilizing an AI tool, an AI solution, or a, you know, a a tool that has a AI embedded, and all of a sudden, it just hits you.
Like, wow. This is this is powerful. This is not your, you know, clippy from back in the day. This is not just some fad.
Like, this this is powerful.
Steven, over to you.
Yeah. On that topic of kind of the the spectrum of of where things are with, how functional the AI is, I think one of the ends of that is is kind of just it functions as, like, a control f for you of, like, I could have just found that myself. I could have just figured that out, in just about as quick of a time as it spat something out. So having these flags that we have set up for, variance analysis, error checking, and things of that nature that allow us to just start from knowing what needs to be addressed and then being able to take quicker action from that point and having it, you know, in big bright red on my screen saying, like, hey. Look here, is super impactful and is, something that we use just about every single day.
Yeah.
Yeah. It's that's another super powerful part about AI is it literally can be used in so many different not only every day, but in so many different situations every day. Absolutely incredible.
So we have, you know, about fifteen minutes less left. I see a few questions here I wanna get to. But before we go there, just kinda wanna summarize, you know, you know, some of the things I'm hearing about, you know, how AI can save time, reduce costs, etcetera.
Any any last kind of, you know, red flags or risks or concerns, you know, from from either you two that, you know, are worth highlighting?
I mean, one red flag would be a resistance to embracing AI. Oh, yeah. Right? Those that those that avoid using it or see it as just a risk are going to miss out on a huge opportunity.
But I do think we have to still maintain data proper data practices, data security practices.
I think we have to guide employees around proper use of AI and improper use of of AI.
And the the, security and compliance functions within companies can help ensure that employees know how to do that.
I think we have to be So there's so there's a both there's a risk of not using it and not experimenting and and getting left behind, but there's a there's a risk of of not using it responsibly and and then exposing the company to risk.
K? Yep.
Yeah. And then there's also a risk of just using it in a novel way. So my teams are making sure it's actually saving them time. Lot of times, AI is doing things behind the scenes that people don't even realize, but the end product is just much faster and the uptake is is much faster. But don't just accept what AI produces and hand it right over because then you may find that it's written something that is not exactly what you wanted to say or it could produce spurious results.
We've all encountered this when we just asked chat GPT something simple and it gives you a wacky answer. The same thing could happen with these tools as as various suppliers try to surface AI in context in the tool, it's still only as good as the learning process has been for whatever that AI model looks like. So I'm not saying fully caveat emptor, like, let the buyer beware, but make sure you're not just accepting whatever AI says as the truth and running with it. Like, validate it. Gut check it.
Yeah. That's really interesting. You know, as a procurement guy, I'll get sent over, you know you know, order forms, and there's you know, they're quite complex sometimes. And, of course, I'll I'll review it quickly, but it's it's always interesting to upload that to a a secure, you know, AI platform and then say say, hey.
Like, kinda summarize some of the risks that you see here. And I'm not gonna take that at face value, but it is a good area for me to say, oh, I'm gonna go double click on this and double click on that. Yeah. Very awesome.
Steven, any any any parting thoughts from you from a from a risk concern perspective or or anything else that we haven't clicked on that that you think is worth sharing?
Yeah. Definitely. I like the concept of the risk.
It's there's a lot of solutions out there that are touting their new AI features, and it's hard to determine who's actually not using your data to train their models, who's actually, maintaining strict data security standards. So anytime we get a request to work with a vendor that has an integrated AI solution or it's just a stand alone one, I'm always having a conversation with our IT manager to make sure that we've reviewed, what needs to be reviewed to confirm that we are not going to be putting ourselves at risk by utilizing this solution.
Because what we definitely don't want is for somebody to copy and paste a massive amount of data into chat GPT to get it to summarize it.
And then now all of our sensitive information is just out there in the e Yeah.
And and, obviously, you know, hopefully, you're checking with, you know, internally.
I I assume companies are enabling, but that's where I really like this concept of AI within, you know, the tool itself where you can do that in a safe, you know, secure place. You know, it's you know, you know, everyone in this call obviously is, you know, familiar with Tropic and and, you know, how we, you know, support AI from a a spend management perspective. But it's been awesome, you know, to kinda see that. And if anyone on the call is interested, obviously, we'd love to show that. But, you know, whether it's Tropic or another, you know, tool, I think the important thing is to, you know, just start exploring how AI can fit into your strategy, but also understand, you know, how you can do that, you know, you know, safely and securely. So feel free to connect with us for for more examples or or a deeper conversation.
With that, let's let's jump into some some q and a. If anyone listening has some, you know, questions they wanna put in the chat, that's great.
Let's see here. The first, first question I'm seeing is let's see here. Steven, maybe we'll start with you.
They said I'm getting pressure to simply use more AI.
How can I help non finance stakeholders understand where finance should start with AI, since it's a little different from, say, engineering or marketing?
Yeah. So I think that, starting small is a very useful strategy because, it's easy to, you know, just sort of dip your toe in the water, gain confidence with the solution, prove the value that you're receiving from it, and that provides a more concrete argument for you to take to your company or for the company to take to you, to approve the request, to press the pedal down even more on it. And, you know, once they start seeing the value and understanding it with concrete examples from, recent past at the company, then, I think that's kind of your entry point to being able to utilize it more, or just finding novel ways that, you know, maybe it wasn't designed to be used this way, but you can utilize it in a whole host of ways that, maybe people haven't really thought of, and you can get additional value out of it that, you weren't expecting.
Like, we really pushed the Tropic system to its limits earlier this year by utilizing a software request form for every single type of, spend request that we have at our company. And now you guys have the general workflows, and there's so many things that are making it a lot easier. But we saw the value in the platform before the feature was actually there. And I think that that can be, applied to, AI tools as well.
Yeah. Very interesting.
Next question, Russell, we'll go to you. It says, as a I'm gonna summarize a little bit. It's a little long. But as a finance leader, I'm often being asked to approve additional funds for both existing and new AI tools. How should I be looking at that and and and not only enabling the team, but pushing back so it doesn't destroy budgets?
Well, first of all, it sounds like, we're requesting to spend money on something, so you you need a spend management platform.
I can think of a good one called Tropic. But assuming you have such a system, I think you have to first think about what is the use case of AI in that scenario.
Because, again, like we said before, people aren't going out and buying AI tools.
They're buying tools that are enabled or empowered or AI forward.
Right? And so what are we hiring this tool or this investment to do?
Is it to identify and ingest, classify, recommend, predict, act, or nudge? And do we agree with that use of AI? And is it duplicative to something else that we're already doing? I mean, it goes back to our spend management principles of which AI doesn't change that. Just because you throw AI in the name of something doesn't mean that all bets are off. We should just throw money at it. But we should figure out, is the AI that's being used in the tool being requested, is it accomplishing the thing that we need it to accomplish?
Yeah.
Which could be summarizing information, saving time, improving our predictive ability, consolidating workflow. Like, what are we hiring AI to do in context?
I have to imagine maybe either one of you can validate that, but, you know, because, you know, in both of your roles, but there I'm assuming there is that influx of of tools and and desire to spend money. Is that is that true?
Yes.
A hundred percent. I've looked at two just this week.
Two a week ago. Used to do that, Steven?
I think it was called Tropic, actually. Funny enough.
Yeah. And it I mean, it makes sense. Right? Everybody's scrambling and excited to leverage the newest, latest, greatest thing, and there are you know, I'm being kinda silly about people throwing AI on the solution name, but there's some really cool stuff.
The the pace of innovation in this category is truly mind blowing, and so a lot of it is worthy of investment. And so now we have this new fascinating, but heavy responsibility as CFOs and finance leaders across the world to figure out which ones are the right ones to make a bet on and which ones are the wrong ones.
Yeah. We have to help the company figure that out.
Awesome. Couple more here. I know we only have a few minutes left, so we'll get them to quickly.
And we've talked about this one a little bit, but what specific use cases and either one of you can answer this. But what specific use cases of AI have you seen provide the highest ROI in finance and procurement? So I guess I could weigh in too. But either one of you wanna go?
And we've shared some examples, so it might be a little bit repetitive. But if if something else comes to mind, by all means.
Yeah. I think for me, the most, relevant is something I already touched on earlier of being able to actually compare and contrast, these potential vendors that we want to, work with and having the information so readily available there to be able to actually make more informed decisions.
And, like, that will pay dividends in the end because doing more of the legwork upfront will help us prevent switching solutions after six months because we weren't getting what we needed out of it. And, you know, maybe it doesn't play as well with our integrations as we want or maybe you know, there's a whole host of reasons that that tool can help us identify as potential reasons to maybe go with a different vendor. And even if that vendor may be slightly more expensive, just the cost of not having to switch in the future from both time to switch and, the cost of actually switching itself.
I think that, to me, one of the most impactful areas for, ROI from the AI tool.
Yeah. I would say, you know, tools out there and financial teams are solving for what happened.
Accounting teams do that really well, and there's some time saving opportunities there.
What will happen? That's the world of FP and A. That starts to get me really interested and excited where you're talking about prediction because I think AI does really well there. It's only as good as what you're feeding into it. And what what could happen?
Scenario analysis. Vectors you didn't even consider. So the the the one where I think the the the strongest ROI right now is in the predictive space.
AI surfacing those root cause analysis or those vectors of consideration that would take longer to consider that your team just doesn't have the bandwidth or capacity to consider. That's that's what has my attention.
Yeah. And maybe I'll just weigh in on this one as well. You know, I I come from a world where, you know and and this goes back to what the both the two of you were talking about the very, very beginning where limited resources well, that's that's always been the case, honestly, for procurement and tech where it's it's kind of a new, function in general. You know, when I joined tech back in twenty sixteen, I didn't have many peers in the procurement world, in the tech community.
Right? And so you're covering a very broad amount of of spend, a very diverse amount of spend, and it's hard impossible in a lot of ways to to be a specialist. Whereas when I was in the world of direct procurement for a big company and one of many, I was very specialized in what I do. And so, you know, AI has the ability to once again help me understand quickly what a supplier does, what their strengths are, what the competitors are.
And then if you and if you're leveraging it within a tool like Tropic, then no benchmarking. And and it and it just empowers me because it saves me all that time for good and up to speed. It makes me so much more effective in what I do in terms of both getting results, but then also just being able to speak the language and understand what it is I'm buying, and that has been an absolute game changer. So alright, gentlemen.
It is, we're we're at time. It has been an awesome conversation. Sincerely appreciate both of you kinda weighing in here. It would be so interesting to do this again in, I would say, a year, but you could probably do it again in six months and have, you know, so many more, you know, insights because it's moving at such a fast pace.
So thank you both. Thank you, Russell. Thank you, Steven. Appreciate you, and, we will talk soon.
Thanks, everybody.
Thank you.
