Adam Larson: 00:21

Welcome back to Count Me In. I'm your host, Adam Larson, and today I'm excited to be joined by Bill Koefoed, Chief Financial Officer at OneStream. We're digging into the rapidly evolving world of AI and finance, especially how it's revolutionizing financial forecasting and reshaping what it means to work in accounting and finance today. Bill shares his perspective on early adoption of AI, why transparency matters, and how these tools can actually make finance roles more engaging and strategic. We also touch on the essential skills the next generation of finance professionals will need and how leaders can foster a culture of innovation.

Adam Larson: 00:51

Let's get started. Well, Bill, we're very excited to have you on the podcast today as we're going be talking about AI and financial forecasting. And I figured a good place for for us to start would be how is AI currently being leveraged in finance, in stuff like demand forecasting?

Bill Koefoed: 01:14

Yeah, I mean, demand forecasting is a great use case for particularly for machine learning, the more quantitative aspects of AI. You know, in today's uncertain world, people need more information. They need to be able to do more scenario planning. And in Canada, the stakes are as high as it's ever been, because if you miss a forecast, either too high or too low, you know, it's real money. And we found a number of our customers have been receiving better forecasts, and they've been doing it in substantially less time leveraging AI.

Bill Koefoed: 01:49

And and that's really a great opportunity, and and I think we're just at the early stages of that.

Adam Larson: 01:54

So we're, you know, we're at the early stages at it, and and do you think that it'll be prohibitive of people adopting it? Because, you know, sometimes when something's in an early stage, especially of a technology, they're not so keen on trusting it, especially something like an artificial intelligence or like, you know, should I put all my eggs in one basket kind of thing?

Bill Koefoed: 02:11

Yeah. I mean, I'd say two things about that. One is there's a lot of hype around AI right now generally. And I think that there's been some important use cases there. You know, I mentioned one obviously using things like generative AI for helping developers do a better job in terms of writing code has been a pretty good ROI associated with it.

Bill Koefoed: 02:37

And clearly, call centers have also been, I think that path is well worn, and certainly it is within our company. But I that it, I think to your point, kind of begs the question of like, how transparent are you in the AI, you know, inputs and outputs? It's one of the things that we've really focused on is rather than just having a black box like some companies have, which really scares data scientists because they can't trust. You press a button and out comes a number and you're like, I don't know how that came about. We really exposed the inputs and the outputs.

Bill Koefoed: 03:16

And for us, we basically run this concept of a model arena where we let different models kind of compete against each other such that you get the best answer. And that could change obviously over time. But we give all that transparency to our customers, and they want that. If you don't have that transparency to your point, you're going to really struggle to get adoption of your AI capabilities.

Adam Larson: 03:41

So what are some other examples of, you know, how AI has kind of made a significant impact? You know, things like maybe like allocation of resources with an organization or other elements that you've seen?

Bill Koefoed: 03:51

Well, I mean, mentioned a couple obviously with our developers and and with with contact centers. I mean, I think you're going to see it even more broadly utilized as I think people get more comfortable with use cases. There's another dynamic, obviously, is that you don't want proprietary information kind of domain. The And so making sure that you've got, you know, like OneStream is a book of record for most of our customers, if not all of them, and you don't want, this is the information that's going to Wall Street. So you don't want that obviously out in the public domain.

Bill Koefoed: 04:26

So how do you actually take in LLM models, use them within your environment and then be able to look at, as a finance person, today, I'd love to look at revenue by geography, how much marketing spend I've given, and I want to do it by period. I can go do that today, but I'm going to have to have some analysts that are basically kind of wrangling and untangling data to figure that out. Whereas, you know, this is going to be a perfect use case of saying like, Hey, I just want my AI agent to be able to tell me that, and I want to be able to do that in a fast time. The other thing is obviously, I think we're all looking at whether the data is actually accurate. And I think that's going to improve over time.

Bill Koefoed: 05:12

And so I think there's to some interesting use cases. I get asked sometimes, you know, how's it going to impact finance employment as an example. And I think it's actually going to really supercharge finance. Some 22 year old that's graduating from college doesn't want to spend all their time just sifting through data, putting it in the spreadsheet. They want to use their brain.

Bill Koefoed: 05:33

And this is going to give them a lot more of an opportunity to get all the information that they need to be able to make better decisions. And I think that's going to actually be really exciting for the kind of accounting and finance profession.

Adam Larson: 05:46

Well, yeah, especially as, you know, when when, you know, twenty years ago when somebody's coming out of college going into the finance profession, a lot of things that those people were doing are handled by things like AI now. So it's going to change the landscape of accounting education.

Bill Koefoed: 06:00

Well, for sure. And I think part of the issue that faces the profession today is that there's less people going into the profession because they don't want to, you know, you can decide, want to go into accounting, or I want to go into investment banking, or I want to go into law or some other profession. And I think part of that was just because they felt in the accounting profession that they weren't, you know, that it was just a bunch of, you know, number crunching, shall we say. Whereas I think this is going to give them an ability to have a much more enriching job and one where they can add a lot more value the company that they work for. And hopefully that'll turn the tide a little bit in terms of more people going into the profession, but certainly it's a challenge that we've been facing here over the last few years.

Adam Larson: 06:46

So for people who are in the profession, you know, how can what how can they upscale to to be at the cutting edge? Because, you know, obviously what we were doing before isn't going to work. You know, what we're doing in the future, technology is constantly changing and they don't want to fall behind.

Bill Koefoed: 06:59

Yeah, it's a great question. I think that, you know, being, you know, having an insatiable curiosity and really embracing AI rather than being kind of nervous about it, I think would be the advice that I would have. Go learn more about it. Try it. Go on something like ChatGPT and try different solutions and questions and just get more familiar with how it works.

Bill Koefoed: 07:28

Because there are some nuances in how you ask the question and what kind of data you provide all that. Clearly there's application. Ended up using The first time I used it, I used it for writing a job description for a new hire that I was doing. But, you know, the more you get more comfortable with it, you know, Microsoft obviously has Copilot that they make pretty easily available and use it and try it and explore. As you do more, then you'll get more comfortable with it and you'll get better at it.

Adam Larson: 07:56

So you mentioned, you know, you've been using it in your daily life to help with things like writing job descriptions. I know people help with their emails. You know, are there other ways that you've seen, especially people within the finance and accounting function, utilize it for day to day tasks, maybe even help just make their jobs more efficient?

Bill Koefoed: 08:14

Yeah. Well, certainly, it kind of borders between kind of legal and finance, but being able to go look in contracts for specific amount, for specific amounts and or information, if you can put those all into your LLM models, it'll help give you more insights. You can certainly do the similar process and look at trends that you have around pricing and or packaging and or demand. I mean, there's so many opportunities for finance people to be able to use it just to get more insights into their customer trends. In addition to that, things like account reconciliation and transaction matching are clearly an opportunity to better utilize, again, a little bit more on the quantitative side, to be able to use AI and, and again, make their jobs more efficient, more effective, more useful.

Bill Koefoed: 09:11

The list goes on and I probably haven't even contemplated other opportunities for how people are using them, but it's a really exciting time. And, you know, as I said, I think it's going to be enriching for the profession.

Adam Larson: 09:27

Now, I wanted to touch on something you mentioned earlier about, transparency. You were saying how you guys make sure you're transparent and open with how the AI is being used and the assessment there. How important is it for organizations to have their own policies around AI, especially, you know, with organizations you work with?

Bill Koefoed: 09:43

I think it's really important, again, particularly, I mean, your financial information is kind of the keys to your kingdom, right? I mean, it's the most proprietary, or one of the most proprietary assets that you have, especially if you're a public company. And so, you know, we obviously take security very seriously. And so I think it's, I think it's really important that as you start thinking about, you know, your, your, the AI tools that you want to use, you do it in the context of your security environment and make sure that it's consistent with that, environment and that you're protecting your information.

Adam Larson: 10:21

Yeah, definitely. Because it's it's one of those things where anybody with a connection to an Internet can access an AI application. And what are you putting into that? What does that AI application have access to, you know, and how are you utilizing it? And I suppose that's why a lot of folks when they're working with, software, you know, different softwares that they're working with, a lot of times those softwares have included some sort of an LLM within their platform to help, you know, utilize the functionality just within a safe environment.

Bill Koefoed: 10:50

No, I think that's right. Mean, and I think, you know, whether it's us or ServiceNow or some of the other, you know, big software companies, I think you're seeing that trend pretty consistently applied that you're using AI within the context of the software provider.

Adam Larson: 11:07

Have you guys run into any, like, around the conversations around, you know, ethics within the AI platforms? You know, it's different types of biases that can get introduced when when looking at different number forms. How have you guys addressed those kind of problems?

Bill Koefoed: 11:21

Yeah, we've been a little bit more oriented on quantitative AI than some of those issues, shall we say. And certainly on the LLM side, I think they exist a little bit more. But I think, you know, our sweet spot, given that we are a financial software company and particularly around numbers, you know, again, there clearly can be errors if if, you know, we we tend to be more focused on the quantitative side. Okay.

Adam Larson: 11:50

Now, how important is it to have the human side of an A. I? You know, you have the A. I, but a lot of times people say you also need the H. The human intelligence, to to kind of come together.

Adam Larson: 12:01

You know, how have you as an organization been able to, you know, utilize the human side to make sure that the A. I. Is doing what it needs to do? And then also, you know, because, you know, we want to trust the numbers that are coming out. But obviously at some point, somebody has to check things to make sure things are still on the right track.

Bill Koefoed: 12:15

Well, I mentioned a couple of these things previously. I mean, I think transparency really matters so that you can get people to trust your data. Again, to the extent that it's it's accessing and finding data and being able to expose that data to humans, I think, again, that saves people a ton of time. I mean, just using, you know, on the developer side, you know, we have a lot of developers that are using it on the LLM side. I know I'm going back and forth a little bit between AI and ML, but on the developer side, like they love it.

Bill Koefoed: 12:49

It makes them more productive. They have to spend less time writing code and they can leverage code that's been done before. I think it gets a little bit back to my earlier comment about just trust and we love the word supercharging, but I think at least in today's world, I think AI has a really great opportunity to help supercharge humans. And I think that, like I said earlier, I think that'll enrich people's work, their lives, and again, the productivity of the people that are leveraging it.

Adam Larson: 13:24

Do you think that as we look at how AI is being interspersed within, you know, especially the finance and accounting community, do you think that the lack that the the lack of people coming into it, less people are coming into those jobs? Do you think the fact that AI is helping will take in it will take account for that for that less people coming into the industry?

Bill Koefoed: 13:45

Well, certainly it will make people more productive. And so I think that will be quite a good benefit or positive benefit. I also think that, again, to the extent that it makes people's jobs more enriching, I think you could find more people entering the profession and you could see that trend ticking up, you know, from where it's been. I mean, from the days when I graduated from college, I started my career in public accounting. You know, I reference, sometimes investors ask me like, you know, why is consolidation hard?

Bill Koefoed: 14:17

And back when I used to do it, I used to have this, you know, yellow paper and you used to have a column, which is entity A, and then you'd have a column that's entity B, and you had to eliminate all the transactions between A and B, and you had to do it all on yellow paper. I had this, for big companies, I had this yellow paper all taped together, but in today's world, don't have to do that. And so again, can leverage your brain more and you can be more impactful in the work that you've done. We did a study called Finance 2035 that we released, I don't know, four or five months ago. One of the things that we found is that, particularly for public companies, that people expect the CFO and his or her organization to be more strategic over the next decade.

Bill Koefoed: 15:08

Now, the role of the CFO being a backward looking reporter of information, those days are over. You're seeing the CFO role helping a CEO look around corners, and the data that we found from our Finance 2035 study really showcased that, that this is something that CEOs are expecting from CFOs, that investors are expecting from CFOs, that, you know, owners are expecting from CFOs, and I think that AI just really helps accelerate that trend because, you know, the CFO is really an important role and can help, you know, really impact shareholder value for the companies that they they run.

Adam Larson: 15:54

Yeah. Well, we've seen that that trend, especially in the CF. The role of CFO is constantly changing and it's evolving even more that that business partner, that person with the seat at the table who's not just coming with a bunch of spreadsheet was what was actually coming with no information. And it's it's super important. How have you been able, as a CFO, been able to help your team and guide your team with all these constant changes, especially within technology?

Bill Koefoed: 16:19

You know, you have to you really have to embrace it and you have to you know, I'm a I'm a I love to ski and I'm very much of an advocate of like, if you don't fall down, you're not trying hard enough. You're not really trying to improve. You know, you're not trying to improve. Like I can go through a whole day and not fall, but then I do try I try not to fall. But you have to push your capabilities.

Bill Koefoed: 16:45

I think as a finance organization, it's one of these funny organizations that you have to be right. I mean, the numbers have to be right. They have to be right when you report them to Wall Street. But the process of getting to right doesn't always have to be right in every part of the journey. And you have to be willing to try new things and you have to be willing to explore.

Bill Koefoed: 17:04

We ended up, we use Microsoft as our tech platform for one stream. You know, I take all my Microsoft data and I take all my revenue data, and we've created this process to create gross margin by customer, but we didn't get it right the first time. A variety of different reasons on why we didn't do it, but you just got to keep trying and you got to reward people for innovation and for trying. And at the end of the day, the numbers do need to be right. But the process of getting more insight and making you smarter and better and more strategic, I think is a really important one.

Bill Koefoed: 17:46

And I think it's a culture that, know, as a CFO, you really have to perpetuate. That's how you create great teams. That's how the teams perform. That's how they accelerate. And, you know, at at the end of the day, hopefully, I'm I'm also a big believer that that work should be fun.

Bill Koefoed: 18:08

And I I try to make it that way.

Adam Larson: 18:12

Well, I like what you said, you know, the first time you didn't get it right. And I think a lot of times, especially in the finance function, you don't want to not get things right because if you get something wrong, like let's say you miss a zero, that zero could have huge consequences. So how do you create the space to maybe not get it right the first time, but also, you know, hey, we want to get the numbers correct as well?

Bill Koefoed: 18:32

You know, you really have to create a learning culture. Mhmm. So I I think one of our one of our board members is is General Petraeus who ran, you know, both, I think both the Iraq and Afghan war. He was the four star general that was in charge. He he makes it's it's fun being in board meetings with them, because he's got a lot of he's got a lot of really interesting phrases, but one of them is like team who learns the fastest wins, because you don't always go into a situation knowing the answer, or even the answer that you think you know when you start may change, but you have to learn and you have to learn the fastest.

Bill Koefoed: 19:15

And I was really lucky. I worked for Microsoft for ten years and I think that was part of their culture was fail fast and, you know, and learn from it and and iterate. And I think as a CFO, you have to, you know, you have to do that because we have to learn. We have to get smarter. And the more strategic you are, you've gotta you've gotta really embrace that.

Adam Larson: 19:38

Yeah. You really do. And I like that I like that mindset, and I think it could go far for especially for keeping high performing talent because you don't want people to come in and just kinda get bored or get overwhelmed by all the work that they have on them. You want them to excel and to be innovative.

Bill Koefoed: 19:51

For sure. You know, I also if my team's listening to your podcast, they're good to probably start laughing. But, you know, sometimes, you know, when we finish our monthly report, you know, I just kind of, riff on whatever's kind of top of mind so that you can I want everybody in my team, I want everybody in my organization to know what problems I'm trying to solve, what are the things that I'm worried about, Because if you don't do that, then you can't like People can't help you if they don't know what your problems you're trying to solve? And so it's really been a fun part, at least fun for me, fun part of the team to know, to kind of bring everybody along and make sure that everybody's all, you know, all knows at least what's top of my mind. I want to know what's top of their mind as well.

Bill Koefoed: 20:40

Certainly they're not always as communicative as I would hope. I do reach out and try to listen. Sometimes And that's better just one on one that isn't like group setting.

Adam Larson: 20:50

No, I appreciate that. I appreciate that perspective and, you know, I hope that, you know, I hope that they're enjoying it as much as you are, you know, because a lot of times it's reflected in the leadership what, you know, what's happening with the employees as well.

Bill Koefoed: 21:02

You'll get a bunch of comments on your podcast from my team.

Adam Larson: 21:05

So, you know, just bringing the conversation back around, we we've been talking a lot about a lot of different things, especially when it comes to A. I. Within the finance function. You know, are there are there different trends or technologies or something that you kind of find exciting? You're like, hey, I can't wait for this to happen.

Adam Larson: 21:18

This is going to be really cool, especially, you know, when I implement my team or we implement as an organization.

Bill Koefoed: 21:23

You know, I do think that as, you know, we talked a little bit about earlier about some of the kind of strengths and opportunities around the LLM models. I do think that they're going to continue to get better. I don't think, actually, I know that they're going to continue to get better. Of course. And I do think, back to the example that I gave of, I'd love to look at the truck.

Bill Koefoed: 21:43

Mean, every CFO would love to get more insight into, hey. If I do this, what happens? Or I did that, and what happened? Or are there some trends that I can see that I can replicate or things that I that I did that I don't want to do again. And I think that the AI is just going to allow you to do that analysis that much faster.

Bill Koefoed: 22:05

Back to my earlier point about he or she who learns fastest wins. I think that AI is gonna really help you learn better and faster, and it'll by definition, I think the AI is gonna get gonna learn and and, you know, help to provide some recommendations or some suggestions. If I could ask and say like, Hey, if I spent a million dollars more marketing in Europe, how much revenue would that lead to? That would be great. I mean, we can go try to attempt that analysis, but just the more you're better and smarter, I think that'll be I think it'll be an interesting phenomenon.

Adam Larson: 22:46

Yeah. It will be interesting, especially when it comes to things like forecasting where you're like, hey, I wanna look at the forecast, but you can ask the model saying, hey, compared to the last twenty years of data, what is our next six months going look like? It'll probably be much more accurate than what we're guessing based on, you know, everything in our heads.

Bill Koefoed: 23:01

Well, and even beyond that, how can I change the forecast? Right?

Adam Larson: 23:05

What Yeah.

Bill Koefoed: 23:05

What are the levers that I can, that I can turn to change the forecast if I hire more salespeople or if I spend more money on marketing or, you know, I do I spend it in France versus Mexico. All the all those things are are what a CFO's thinking about every day.

Adam Larson: 23:25

Yeah.

Bill Koefoed: 23:26

In terms of, like, not just what it what is the kind of forecast, but, like, how can I change it? How can I influence it? I think that's gonna be really exciting.

Adam Larson: 23:34

Do you think we're gonna see, investors kind of looking for, you know, new IPOs or even people that are investing in and saying, hey. How are you using AI? Is gonna be a question that's gonna start being asked, do think?

Bill Koefoed: 23:46

I mean, we we were certainly I mean, we have an AI product, so we were certainly, asked about it. And, you know, the question is always like, like, how are you using it in your own environment?

Adam Larson: 23:56

Yeah.

Bill Koefoed: 23:57

Because, of course, you know, they they wanna make sure that you're drinking your own champagne or eating your own dog food, whatever term you want to use around that. So I do think that's right. Think investors always want to know kind of what your product vision is. I to some degree, are you using AI to accelerate your product vision, I think would be something that they would clearly be looking at. I think also on the efficiency side, I think that that they, you know, that they wanna know that they're asking it.

Bill Koefoed: 24:34

It's usually it's usually not their first question, but it certainly can come up in the, you know, kinda call it top five.

Adam Larson: 24:40

Yeah. Bill, I just I really appreciate your insight. You're coming on the podcast, sharing all this information, and having this conversation because it's an ever changing product. I think if we have this conversation six months from now, it'll be completely different. And I think that's the exciting thing about new technology is you can kind of watch it and kind of observe it and see because it's been around long enough that people are starting to add it to their software's and add it to their things that people already have access to.

Adam Larson: 25:03

So it'll be interesting to see the landscape change as people start to understand how to use it more efficiently.

Bill Koefoed: 25:09

Look, I think we're in a really fun time for technology companies. It was was, one of my prior jobs, I was speaking to a group of technology investors, I said, you know, it must be horrifying to be an investor because literally companies that existed in, you know, twenty five years ago don't exist today, because they've been acquired or they've been, you know, they don't exist for a variety of different reasons. But it's a really fun industry to work in because it's so dynamic. There's so many changing things. And you got to read the newspaper, read different articles online, and you learn a lot.

Bill Koefoed: 25:48

I I do think that it's energizing as a, you know, as an employee in in the technology industry as I've been for most of my career.

Adam Larson: 25:56

Again, I just wanna thank you so much for coming on the podcast, Bill. It's been great to have you on here.

Bill Koefoed: 26:00

Thank you.

Announcer: 26:03

This has been Count Me In, IMA's podcast, providing you with latest perspectives of thought leaders from the accounting and finance profession. If you like what you heard and you'd like to be counted in for more relevant accounting and finance education, visit IMA's website @www.IMAnet.org.