if you would like to earn CPE credit for listening to the show visit earmarkcpe .com slash fpa download the app take a short quiz and get your CPE certificate finally if you enjoy listening to FPNA today please go to your podcast platform of choice click the subscribe button and leave a rating and review of the show.
And now onto the show.
From DataReels, this is FPNA Today.
Welcome to FPNA Today.
I'm your host, Glenn Hopper.
Today, we're joined by Dave Sackett, a forward -thinking finance executive and co -founder of AI One, where he's pioneering AI -powered e -commerce.
With deep experience as a CFO, ERP system architect, and a vocal advocate for blockchain and AI and finance, Dave brings a unique blend of tech savvy operational rigor and a genuine servant leadership mindset.
He's a frequent keynote speaker, a Forbes contributor, and an active member of Financial Executives International.
In this episode, we explore how modern CFOs can embrace AI, rethink forecasting, and build finance teams ready for tomorrow.
Dave, welcome to the show.
Hey, thanks, Glenn.
It's been a while since our last webinar together, and for those who haven't heard you speak and aren't familiar, could you maybe give us a quick version of your journey, how you got started in finance and what you're doing today?
Sure, absolutely. So my background is cost accounting.
So, decided to be a cost accountant to let people know where are you making money, where are you not making money.
And from there, you kind of rose up to the CFO level, working 20 years for a Japanese -owned semiconductor company.
About 2016, I got really interested in AI and started talking about AI, started telling people about what's coming, and happy to say that, yeah, that forecast of where AI was going is coming true, and it's quite amazing today.
And you've described your leadership style as grounded in a growth mindset and servant leadership.
How has that influenced your transition from kind of individual contributor to CFO and from CFO to tech entrepreneur at this point?
Yeah, one of my strategies has been to adopt servant leadership.
Meaning I give my team resources.
That my losses have always been the traditional, do what I say, my idea is the best. And I wanted to not be that person.
I wanted to listen to team members.
I wanted ideas not just to come from me, but from my team, and really empower them to have success, so that they had good job satisfaction, and they could have their own wins, to have their careers advance.
You know, I'm already at the top at CFO. I don't plan on going to CEO, you know, my job now is to manage the team and make sure they perform in finance so that the whole department and responsibilities works well.
you and i both talked previously about being introverts and you know now you're out there speaking at conferences you're writing for forbs you're doing webinars you're really out there as a thought leader and i'm wondering as an introvert how do you make that shift do you have to psych yourself up for it and and what's what's kind of your approach to this as an introvert okay yeah that's an and that's an excellent question i used to be terrified of public speaking um in school i remember counting who was going to be up next so that hopefully i didn't have to talk that day and just absolutely afraid
of what people are gonna think of me, based on my opinions, what I said, how I said it, if I stumbled on my words, if I used the filler words, like, I'm terrified.
And in my mind, so fear and excitement are very close if you study brain technology.
So it's really having that fear and turning it into excitement.
And then in my mind, I'm someone that's trying to share my information and my knowledge, so I'm trying to help people.
So in my mind, I'm here talking not to do anything else but just help people in the audience understand my experience and where I think things are, information I should share.
So it's definitely from fear to excitement when it comes to public speaking or writing or putting my thoughts and ideas out there.
So one of your recent webinars that I saw, you talked about the costly failures of traditional forecasting.
and I think that that's a very timely webinar right now.
And I'm wondering from that, and from your experience, what's the biggest kind of mindset shift that FP &A teams need to make to modernize their approach, and what does an AI -augmented forecast actually look like in practice?
Yep, so I would think, come at it from your end product, what are you gonna deliver in terms of forecasting?
How is that gonna, can you back up that forecast?
What are the assumptions that go into that forecast?
You know, traditionally you're using internal drivers like your backlog, new product releases, you know, things very traditional to FP &A to do a forecast. But my prediction is that you're gonna bring in outside influences and outside drivers external to the business and blend the two so that you've got a dynamic model that predicts the future and how well you train your model and add different things into it, you're gonna get better forecasting results and hopefully, that's gonna protect your company by having a good forecast. Yeah and it feels like we're at this moment of economic uncertainty
again. It's not unlike the global financial crisis but with different pressures this time.
And I think back to all the different scenarios we were trying to model back then and the tools we had.
So this is 07, 08, 09, very different world technology wise than what we're dealing with today.
So based on the tools that are out there now, do you think AI enables a better approach to scenario analysis than what we had back then?
Yes, so I'll even...
I guess I can bring in AI at this point, because using AI and forecasting, you can do a lot of what -if analysis and just change a few parameters, press a button.
Bang. Now I've got a new forecast, and that doesn't disappear.
Now you have it, and if you're training your model and training your bot to be accurate by doing backtesting and, including all the factors, reading your ERP data right, you know, you can use that as instead of number crunching and having a junior FP &A person just put everything together, you're having AI do that automatically.
And you know what's driving that forecast, what were the assumptions at the time, what was the backlog, what was the new product release schedule, the timing of sales, you know, all of that can be documented and kept by AI.
So that now the team isn't just putting numbers together, they're guiding the models and they're actually influencing how AI is going to be useful as a tool.
So let's talk practical use.
I mean, if you're a giant enterprise out there and you've got a data science team, you're probably doing some really cool stuff in the company.
But for the rest of the world, mid cap, SMBs, using generative AI typically means taking data out of your workflow, uploading a CSV in the chat GPT, working with it, and it can do amazing stuff, but it's a little bit clunky.
And we know things are changing fast, but what do you see current best practices and where maybe FPNA software goes in the next three to five years, say?
So from my prediction point of view And where I think we're going is that every FPNA software is going to have an AI component.
Whether you realize it or not, it's going to be built in and people won't have to become data scientists in finance.
They'll just have to be able to use the software as a tool just as you would use a calculator 30 years ago.
Now it's going to be, hey, your skill is going to be using AI within software.
So I do agree, most companies in, you know, probably on the planet are SMBs.
Even though the big companies get all the press, but most people are in smaller companies that can leverage AI, it's just not, it won't be through an SAP, you know, design to have it built in and have, you know, in your menus to choose and to run, and it'll be very much far low -key.
And just, you know, company to company, It's like where do they need the help with AI?
And that's where they'll adopt.
I always say that people in finance don't need to become data scientists.
But we do need to understand the fundamentals of data science so we can ask better questions.
And the example I always give is a simple one.
If you're anyone who's in management and has been for a while, you've seen financial statements, you can read them, you understand this is what goes on the balance sheet, this is income statement, this is cash flow statement.
But maybe they're subtleties that you don't get, Like you don't you maybe looking at operating income, net income, EBITDA, not being a domain expert there that could be an area where, you know, you need to have that domain expertise.
And similarly, it used to be, if you were going to be a data scientist, you would have to learn programming.
So you could understand the fundamentals, but the barrier to entry was, oh, and by the way, you have to learn how to code in Python. So that kept a lot of people from being data scientists.
But now your naturally spoken language becomes sort of the new programming language so you can access data science tools without having to write Python. But if you have an understanding of those, like this is what machine learning does, it does regression or prediction and classification and clustering and these are the types of questions to ask and this is what a correlation matrix is and this is how we assess how well our models predicting So if you have the basic understanding, you know the right questions to ask.
And I really feel like whether it's finance or whatever department you're in, having that data science skill is gonna help you sort of keep up with the future of the technology.
And I'm wondering from your perspective, what skills should FBNA teams be looking at right now to sort of prepare for that AI driven finance function that's coming?
Yeah, so you do bring a point where you need to know how to use AI.
that's going to be – you don't need to be a data scientist, but you do really need to understand where is it strong, and where is it weak?
Where are your skills as the human going to be important in an AI -driven finance department?
You still need to be accountable for the numbers.
You need to know how AI arrives at a certain value.
How can you prove out what AI did was correct?
You don't want to blindly accept the AI's output, and at same time AI shouldn't blindly accept a human's input.
So I think there's going to be that symbiosis in the future where anything an AI does a human checks anything you know a human does AI checks and it's going to be that kind of internal control check and balance within the department so that you know everyone's on the same page and you've got back up in terms of what the numbers really mean and what's really driving your business results.
Yeah and you said you said something really insightful there, that not only should humans validate what AI produces, that human in the loop that we talk about all the time, but that AI should also validate what humans produce.
And I think that's not a framing I've heard before.
And I think it's so smart.
And I think it's recently, I think the timing is interesting too, because these models learn from human feedback and so like chatGPT and Gemini and and and anthropic all out there, they'll give you responses and ask you to rate the response.
And they'd give side -by -side responses.
And I guess, um, ChatGPT was giving responses, and the humans were picking ones that were very friendly to them and maybe have a little bit of that sycophantic.
And it was maybe on these responses, it was subtle, but, uh, it's the language model telling you, the chat bot telling you, oh, yeah, that's, that's brilliant, you're right.
And people were clicking those with thumbs up a lot more than just the regular sort of neutral responses.
So when ChatGPT rolled out a new model a couple of weeks ago, they really they overrated these sort of sycophantic responses.
So it got to be, we know, we talk about bias in AI sometimes, but these models were so tuned to be like telling the user, each user that they're great, that it wasn't doing that.
It wasn't checking them, it would, you could tell it the worst business idea in the world.
And it would come back and say, oh, that's great.
You're brilliant, and all that, but that was an error in training and it was so bad that OpenAI actually had to roll that model back and go make the corrections to it.
So, I think that that's a very interesting concept, the idea that we're checking the model's work and that the model is checking ours and it also shows something that we have to solve for right now.
Yeah, no, I agree. And it's the dopamine reaction that people want to see the little notifications and the thumbs up.
And that's a human thing that we're doing, but do you want AI doing that?
it's more, you know, I'd much rather have an objective.
Here's what the data says.
You know, from the human point of view, I'll be able to take that into a narrative in terms of, you know, should we do this?
Should we not do that?
What actions do we take based on this objective data?
Let's talk tools. How are you and your team actually using AI today, using ChatGPT or more traditional tools?
What are you what are you guys doing today?
Yeah, you know, I mean, from a personal point of view, You know, people are definitely using chat GPT, you know, co pilot all the tools that are out there to help write business documents, make an email sound better.
Yeah. There's you know, people are kind of dipping their toes in the water from a business point of view.
I'm doing more of like robotic process automation that was popular years ago.
But even now, now it's, you know, a cloud app.
It's able to, you know, interact with my department.
And so they're using AI in that, that way now, at least at my company, in my experience.
So I think there's many different directions to kind of get your feet wet in terms of using free tools, having, you know, corporate tools to really kind of future proof yourself, like you said earlier, and just be aware of what it can do and what it can't do.
And if you can find solutions that are non AI, I'm all for it.
But if you do find something where AI is definitely the best tool for the job, let's take a look.
Yeah and RPA was for years, that was the state of the art.
That was when you talked about actual practical use cases, what was being done in finance, RPA was it.
Now we're seeing that that's generative AI is really taking a place there.
Even the RPA companies, UiPath and all are leaning into agents and to using generative AI in RPA.
And I think about this, they're not there now, and I know there's been a lot of talk about agents, but, like when I see operator from OpenAI, it's pretty cool, it's a little buggy, but it's more of a sign of things to come.
So, but when you watch operator work or computer use from an anthropic or any of the other tools out there, it reminds me of RPA.
You see the mouse moving around interacting with the machine the way a human would.
and that made me think of RPA.
But the idea is contextually with generative AI, you wouldn't have to have that level of training that you have to do for RPA, that it just contextually picks it up.
So you see the future there, and it's going to be interesting to see how people, when they imagine if you have your own RPA for bank reconciliations or any basic job that you're doing, that you're able to turn it over to a bot.
I really see that happening.
So for CFOs who want to introduce generative AI into their teams, but maybe they're nervous about privacy data misuse.
What do you like? How do you recommend these finance leaders get started?
What guardrails and governance policies would you recommend?
Yeah, I mean, from a cybersecurity point of view, you know, is it locked down or is a bot going out to the Internet where they're not supposed to be, like really have internal control in place just like they were another finance employee, making sure that what they're pulling is the correct data.
they're linking to the right sources, you're getting the right information.
Humans in finance, like you said it before, we have to have be the, you know, proponents of truth in the data, so people come to us to see what's real.
So we can't get that wrong, and we can't just say well AI did it.
We have to know what did AI do?
Where did they get the data?
Not really a black box where you kind of throw your hands off, but really understand where it's coming, how it's being influenced and how you're getting the information the way you're getting it.
And really keeping that as, you know, we're custodians of truth.
That doesn't change with AI.
That's still, I have to make sure the numbers are correct.
And whether I'm using a calculator, AI, whatever the tool is at the end of the day, the human is responsible to make sure that you're accurate and that it's, you're doing your best with the data you have. INTRO FPNA today is brought to you by DataRails, the world's number one FPNA solution.
DataRails is the artificial intelligence -powered financial planning and analysis platform built for Excel users.
That's right — you can stay in Excel.
But instead of facing hell for every budget, month -end close, or forecast, you can enjoy a paradise of data consolidation, advanced visualization, reporting, and AI capabilities, plus game -changing insights giving you instant answers and your story created in seconds.
Find out why more than a thousand finance teams use DataRails to uncover their company's real story.
Don't replace Excel, embrace Excel.
Learn more at datareals .com.
Another similarity that you and I have and I think it probably influenced the way that both of us think about the role of CFO but you are both a CFO and an ERP system administrator And I think, you know, maybe in the past, that was rare, but there is so much more crossover today between the CFOs role and then our, how important our finance tech stack has become to that.
But it makes you, if you are a CIS admin and you're really leaning that much into the technology side, I would have to imagine that that dual role would change your perspective on everything from data, data integrity, to the kind of the role of ERPs in modern finance.
Yep. Just the ERP keeps all data.
To me as a finance person, I use all data.
Even though I see the output, I want to make sure transactions are correct.
In operations, I want to make sure that the numbers are correct, that there's backup for things, that things are going to the right accounts.
So taking it from really just focused on finance, but now focused it on the entire company?
And are we keeping good data?
And especially with AI and automation, you don't want to start that with bad data.
So you want to make sure everyone's on the same page of whatever that is put into ERP, that's the truth.
And to say these transactions happened outside the system, like, no, the system is there to keep the transactions in.
So it's kind of teaching the company to really be data mindful And to leverage not just the ERP, but understanding what you do has an impact on the business.
And in the front of finance point of view, what you do I can translate into the bottom line and to a P &L forecast. But really it's making sure that it's all done on the right information.
Yeah and there's talk about moving towards the unbundling of the ERP between all these different SaaS tools is that you could put them together with APIs and have them talk to each other, and you could get the functionality of an ERP but coming from many different sources.
And then if you factor that in with what Satya Nadella said a couple months ago about the entire SaaS industry could be at risk from generative AI, because if you think about SaaS tools, they're wrappers or interfaces that go over these big databases.
And what if through generative AI, you don't need these big complicated UX tools you can just chat with the data or build agentic workflows, or have agents going in and doing the work.
That whole SaaS layer goes away, and it's just about the data that's in the database and how we interact with it.
I just wonder where does the ERP, in your mind, fit in the future going forward?
Yep, no I think ERP becomes your the database of truth, I would say.
And you have to link in all your other databases and all your other tools to really complement each other to have a source of the true data.
So linking ERP to different tools and to push and pull information, I think that's all the future.
So it won't be traditionally, everything's done in ERP.
Now, you're gonna have cloud versions.
You're gonna have this software, that software that you'll pull and push information from to get a better picture of what's happening in your company.
Let's switch gears for a moment.
And I do wanna talk about AI One, because I know as a CFO where your focus is, but I'm wondering what inspired you to build this AI powered?
It's an e -commerce platform, right?
So what inspired you to build that from the ground up?
Yeah, no, thanks for asking about it.
Yeah, it's a startup company, based on tracking information and using tools, AI tools to really drive the company.
So we wanted to be AI kind of first. and you can almost like an experiment.
We don't need to just throw bodies at a company.
Let's use AI tools to really drive a company and make that a foundation of how we go forward. It's intentionally choosing that, we're going to use AI to do marketing, to do tracking of sales, to do payments and referrals.
We want to use the AI tools in a company from the ground up.
you know, we don't want to go traditional and switch over to AI.
We want to say, okay, this is new.
This is something we're just making out of the thin air.
Let's intentionally go into the business as we're going to be an AI -driven business.
And this was, you know, years ago before kind of all the hype happened and now everyone wants to be an AI company.
So I'm trying to kind of just shift it to, you know, we're trying to provide a service to match up buyers and sellers, and also have a concept called co -seller where you and I could sell FPNA software.
And if we just do a referral and influence a sale, my technology recognizes that and rewards people who not only close a sale, but influence it.
And so, the concept was kind of started based on today everyone goes to Amazon or Walmart, these giant companies to get whatever you need.
In the future, we predict people are going to buy from their friends and what their friends think is cool.
So by doing that, the seller is kind of leveraging that network of friends to now make sales.
So it's the idea is that the seller wins because he's selling, the buyer wins, they get something that they're looking for.
And then whoever influences or closes that sale get commission from the seller.
and all that's known ahead of time in networks so that my company is, gets rewarded when we build networks of buyers, sellers, and salespeople.
And now you've worked in both large enterprises and now in the startup space and I'm wondering, looking at it from both sides, how do each of these environments approach AI adoption and maybe what could each learn from each other?
Yeah, now that's a good question.
like, right now we're bootstrapped, so we're trying to do everything without spending money, which for a startup is very difficult, but we have got nothing to lose.
So the idea is that when the customers come in, we get paid.
That allows us to now go in this direction and take on this tool.
If we get more business, then now that'll give us budget to do something else.
Whereas my traditional job, I've got an annual budget, I know how much I can spend you know, it comes out of the business.
I'm not personally affected by it with the startup Yeah, I'm putting in the money like I'm the one I'm using that customer cash to now fund building networks and marketing So do I put it towards salaries?
Do I put it towards marketing do I put it towards whatever?
You know I'm very much in sync with cash flow and how that really runs the business versus my traditional, I'm, you know, I've got a sales department.
I have all these different departments all working together.
With the startup, it's all me and my partner.
So it gets boiled down to you wear many hats and you're trying to just keep, at this point, I'm just trying to keep it going, you know.
Nothing that, I don't get a salary from it.
I don't, you know, make money off it yet.
But I'm just trying to build something cool that will eventually make money once I have enough networks and have enough people getting value from my system.
As finance and accounting people, we're risk averse.
That's part of our job and our nature.
But we've definitely hit the peak of the AI hype cycle.
But I'm wondering for people who have been slow to adopt and are now ready to dip their toe in the water and try it out.
For finance leaders who feel like they're behind, how could they catch up without being overwhelmed by all the noise, the hype, and the technical complexity, and just the different benchmarks and everything that's going on around it right now.
I know you and I watch this stuff every day, but for someone new to it, that's a lot.
Yeah. No, I would suggest really researching, you know, agentic AI or using AI agents, which is kind of the where we are today and where the future's going, and it's understanding the capability.
What can I do well and where can it not do well?
If you're a CFO and you're just trying to get and understanding of what AI can and can't do, I think that's a good spot to start.
It's evolved into that, but now you're looking at, these are the functions that I can rely on.
I can have agents verify the warehouse management system that something shipped.
I can have agents organize what's happening in the company and to actually influence policy.
If I look at a customer and they're overdue, The AI agent can put them on hold and depending on what the policy is, you can make that a very strict process or a very open process.
And it's really digging in at that level, not at the technical AI side, but really the applications.
How is it going to help me in the future?
How can I look at AI as a tool that's just going to make what I do more efficient, faster, with a good audit trail if you're in finance?
Yeah. John, you mentioned agentic AI, where like autonomous AI agents take on roles just like they're another junior analyst. And I'm wondering, in that, you know, looking ahead, this is Get Out Your Crystal Ball.
So what's a what's a bold prediction you'd make about kind of where the role of AI and finance goes in, say, five years and maybe what should these FPA teams start, start doing to prepare for that today?
Yep, no, so I'll say yeah, machine learning is the foundation for all the cool stuff that have today.
So, from five years from now, what's not that popular is having a finance assistant or a junior finance person that's AI.
I think that's going to be something that you hire almost, that you're going to have this financial assistant that's AI -driven, and you're going to give them tasks to do that they execute just like they would a human.
And this is the AI side.
They're crunching part of the finance, whereas the other people will be more analytic and presenting, you know, things that are not just the number crunching, the digging, the flux analysis, creation.
It'll be an AI agent or financial chatbot that'll be popular in five years.
Alright, so we always wrap up with two questions, and the The first one, we get all kinds of interesting, it's always fascinating to me to see what we get.
We've had, I've learned that people are huge opera fans, that they run ultra marathons, all this stuff we learn about them.
Yeah, now that I've set the bar high with that, what's something that not many people know about you, something we couldn't find just on your LinkedIn profile?
Okay, well, yeah, even by saying what other people have done, I'm a karaoke superstar in Japan.
Japanese karaoke, that's awesome.
But it's in a very safe environment.
They have rooms, they have the lyrics on the screen, and you get rated based on how well your English is.
I always get very high marks in Japan relative to my friends who are trying to get the English just right.
All right. Finally, everyone's favorite question.
I'm always hesitant to ask CFOs this because we use Excel differently, maybe, than we did when we were really building models, but we all came up using Excel.
So what is your favorite Excel function and why?
X Lookup. I was an index and match guy, I was V Lookup, H Lookups.
But now X Lookup saves me time.
So if I want to tell you, does Dave know Excel?
Well, he knows X Lookup.
He's in our club, that's my favorite.
I do love Excel, and I've used it to get promoted essentially and rise through the ranks based on what I could do in Excel.
So, I have a very loyal to that program and I use it all the time.
Xlookup kind of makes me look like a superstar being able to link data and pull it in.
And it's just a simple formula.
So, Yeah, I love that.
And it speaks to how mastering Excel just becomes a superpower in finance.
And I had similar path early in my career.
And I always said I was promoted because I was good at Excel and PowerPoint.
Didn't even need the MBA, just have the basic Microsoft skills.
But what I didn't realize at the time was it being able to take that raw data, analyze it and then visualize it in a way that kind of tells a story.
Well, that is what moves you up.
That's kind of what we do in FP &A.
So yeah, I mean the ability to use those tools.
And now I have the same approach with AI though.
It's like, that's great.
Excel got us this far.
Now where are we going to go with the new technology?
And it's not time to be burying our head in the sand and building our moat around Excel or whatever.
So Dave, this was a fantastic conversation.
I guess, for listeners who want to follow your work, learn more about AI1, what's the best way for them to connect with you?
Yeah, I'm a big on LinkedIn.
I answer every message, whether you're trying to sell me something or you're asking me a question.
Yeah, the website is AI1 .VC and it kind of shows our idea and we're really kind of looking.
It's not an idea now, it's a future idea that we're building in the stages now.
So, to me it's interesting.
it's we show our philosophy, we show how we want to give back to people and allow everyone in the circle of the network to earn.
So the idea is that everyone wins in this concept.
Alright Dave, well thank you again, thank you so much for coming on.
Yeah thanks Glenn. One.
One. Three. Four. .
