CFO Futures: Navigating the Real-Time Enterprise

Transcript

CFO Futures: Navigating the Real-Time Enterprise

Patrick (00:03):
The way I would put it right now is that the partnership between the CFO and the CIO has never been closer.

Olya (00:10):
I think accountability will still stay on the human. My rule is no human owner means no go live.

Patrick (00:18):
CFOs are inherently risk averse. We have to be. And so we have to be very prudent when we implement AI and agents.

Olya (00:25):
How do we create enough governance to protect the organisation, but at the same time not to govern away the entrepreneurial spirit?

Gina (00:33):
There's no excuse. The accountability has to stick with the humans. As Harry S. Truman said, one of our dear presidents, the buck stops here. You're the CFO and you're signing the name to the financial statements. That's it.

Tom (00:48):
30 years ago, the role of the CFO was vastly different. While they still looked ahead to build budgets, they acted more like historians, studying the past performance to manage the books, taxes, and audits, but largely kept out of board level email chains. Today, most CFOs at the top of their game barely recognise the environment they started their careers in. They are now less a historian and more a fortune teller, sitting firmly at the right hand of the CEO and guiding big strategic discussions. AI's predictive modelling and big data crunching makes wearing the co-pilot hat for a CFO that much more comfortable as they help navigate their companies through global risks and critical business-wide decisions. But keeping the plane on course during the rapid, sometimes turbulent ascent of AI still requires an adept mind, one that can weave through two opposing forces. One's an old hat, mathematical certainty, the rigid, unforgiving demand for data lineage, auditability, and historical truth.

(02:01):
The other has never been worn on the runway before, predictive agility, the magnetic north pulling finance heads towards a real time enterprise that senses risk before it ever lands on the balance sheet. Sitting right in the middle of that tension is the CFO's new tool, Agentic AI, not just a chatbot summarising last quarter's numbers, but AI agents that can monitor demand, run its own financial statement fluctuations, and hand the CFO a set of options before the meeting is even started. But with that autonomy comes new headaches, legacy infrastructure groaning under the weight of transformation, uncomfortable questions about who's liable when it's an algorithm, not a person, that makes a financial reporting error and a talent market where the technical half-life of a skill has shrunk to just two and a half years. Well, with all of that, welcome to the Next Five Podcast. I'm Tom Parker and today we are asking what it really means to run finance in the age of AI.

(03:13):
In an episode we are calling CFO Futures: Navigating the Real Time Enterprise. Joining me to unpack this are three experts. First up, Patrick Villanova, Chief Financial Officer at BlackLine. Patrick, welcome to the show.

Patrick (03:28):
Pleasure to meet you. Thank you, Tom.

Tom (03:30):
Next, Olia Brasi, EVP of Finance Optimization and Governance at Highgate. Olia, thanks for joining us.

Olya (03:37):
Pleasure to be here.

Tom (03:38):
And finally, Gina Gutsite, senior managing director and global leader of the Office of the CFO Solutions Practise at FTI Consulting. Gina, welcome.

Gina (03:49):
Thank you. My pleasure.

Tom (03:50):
Patrick, let's start with you. As I said at the top of the show, the environment that you as a CFO now operates in is nearly unrecognisable from a few decades ago. Given the rise of agentic AI, and I'm not talking about pop-up chatbots, I'm referring to an agent that can execute core accounting tasks. What does the day-to-day of a CFO now look like? Is it a continuous finance model rather than being bound to a monthly or even quarterly earnings calendar?

Patrick (04:23):
Thanks, Tom. I've been in this industry for about 25 years. I've been a CFO for a couple of those years, but my background is accounting. I started at a big four firm, PwC, and just have a lot of perspective just working with executives for 25 plus years and just seeing how their roles have changed and just even my 11 years here at BlackLine, seeing how dynamically things have changed here. The way I would put it right now is that the partnership between the CFO and the CIO has never been closer. If you're an effective CFO, you should always have had that relationship, but now we're literally linked at the hip talking every day. And why is that? More so, we are now data managers and the business, while maybe the CIO handles the infrastructure, handles the backbone of technology, the accuracy of that data, the speed by which you report on that data, the speed by which you synthesise it so other executives can make decisions and execute the strategy, that falls on the office of the CFO, not just me, but my entire team because we are data experts, we're experts in controls, we're experts in accuracy.

(05:33):
So the business looks to us and says, "You have this proliferation of data and we need the CFO to take that, consolidate it, synthesise it, and package it in a way that we can make effective decisions and understand and comprehend it." Now, what happened in the last couple years? You had the rise of agentic AI and that is game changing to the next level. And you mentioned several things in your opening comments. That's exactly what this company does. It has agents that perform standard accounting tasks.
Now, what does that mean for somebody like me? Well, CFOs are inherently risk averse. We have to be. And so we have to be very prudent when we implement AI and agents, to your point, more than just a chatbot or writing a funny email or fixing a PowerPoint. These are agents that are impacting your underlying financial information that public investors, shareholders depend upon.

(06:27):
So testing it, being prudent, being very careful, that's our new responsibility. And that falls once again on us. That's not the CIO's role. The CFO signs the 10K, the CFO asserts to the accuracy of the financial statements. We own it and it's our job to test it and make sure that it's working exactly like a human being would perform the same task.

Tom (06:50):
Yeah. Olia, Patrick said that it's about to test it to see whether the agent and the software is working exactly like a human would. But from an operations side, autonomy comes with a coordination problem. Picture if you were one agent optimising cash flow by delaying vendor payments while another agent is simultaneously flagging those same delays as vendor risk. Without a shared semantic layer, you end up with agents working at cross purposes and a feedback loop of errors. How real is that risk inside a finance operation like Highgates, for example, and who ends up being the digital conductor, making sure the whole orchestra is playing the same tune?

Olya (07:32):
I think the risk is very real and I would challenge that premise a bit. While technology already exists and you can have organisations where all agents run autonomously, we don't use autonomous agents at Highgate quite yet. And before we do, I think a few things have to be true. Data infrastructure has to be in place. You have to be able to trust the data you're making decisions on, and agents do not fix bad data. They amplify it and they make it move faster. Security layer is extremely important. Our principle of least privilege that we governed our human employees with is becoming more and more important because of how fast things can go bad. Well-defined processes, we must have a blueprint where agents execute exactly what we want them to do. And I think without having that human perspective, "Oh, this looks wrong," agents don't have that.

(08:28):
They'll just execute on the process that you put in front of them. Obviously in accounting and finance, our financial statements must be auditable 100%. Everything has to be supported. Every decision must be documented. So the data retention documentation requirements must be in place. And a couple things that I even haven't thought about before I started doing my research in governance are monitoring the drift of the agent work, figuring out what is the agent learning and how they're changing since their implementation, and really focus on retraction and incident response. What do we need to do when things go wrong? I think accountability will still stay on the human. My rule is no human owner means no go live. I think ultimately the human decisions have to matter. And a lot of times I'm asked how do you balance optimization with the governance? And we operate in a very entrepreneurial business environment and we want our people to experiment with technology and bring the solutions to our teams and our ownership groups and reap the benefits, but we want them to do it with guardrails and do it safely and responsibly.

(09:46):
So making sure that when we talk about, again, production of the financial statements, it's 100% auditable. So the question that we are working through right now is how do we create enough
governance to protect the organisation, but at the same time not to govern away the entrepreneurial spirit?

Tom (10:05):
Gina, let's go a little further. There are warnings about emergent misbehaviour in multi-agent systems, the idea that agents could effectively coordinate their way around internal controls while still appearing policy compliant on paper. Something that is called approval threshold arbitrage. The name in itself doesn't sound that scary, but in a reality it very much is. How much of a threat is this? Can and should CFOs and their teams even start to think about the consequences of this and the solutions to it?

Gina (10:37):
I think part of it is the groundwork that we've all been talking about. In order to make sure that the system you put in place, the agents that you have, there needs to be clear guardrails, clear ownership. You cannot replace human judgement . And so if you think about the warning signs of a multi-agent system, that really just highlights the data and maybe the disruption within the data. If you have the right governance, the right information, it should be easily monitored and managed. For example, we do cash flows all the time. We're using AI agents to do it better, faster, more information comes in, but it doesn't replace the fact that someone needs to explain it to the business partners to be able to look at judgments, to be able to understand where that's coming from. And really the communication is key. That human interaction I think will never be replaced by agents or any other tool.

(11:31):
I think it's more about putting in those checks and balances, those guardrails in order to ensure compliance.

Tom (11:37):
Yeah. And Patrick, Gina said about these guardrails that are put in to create compliance. I'm going to put the same question that I put to Gina to you. Do organisations need to catch this kind of behaviour between agents reactively in real time or does it need to be engineered out before the agency even leave the shop floor as it were?

Patrick (11:56):
Well, given what we do, it's got to be, I like how you phrased that, Tom, engineered out during the testing phase. You can use an agent, as I said before, to make a t-shirt or send an email or do a PowerPoint and you can afford to make mistakes in those areas. You can't in our world. It's something I've said several times on other webcasts, but 95% right is 100% wrong in accounting and finance. And that's factual. The SEC says if your P&L is more than 5% off, in all likelihood you have to restate. That's a material weakness. That has career implications, that has legal implications. You cannot afford for that to happen. So when I hear maybe people outside of my universe say, "Oh, I built this agent and it's 90% of the time it nails it." I'm like, "That's failure. That doesn't work. A 90 might be an A minus in school, but it doesn't work in my world." So testing it before it's released into your financial systems ecosystem is critical.

(12:57):
Engineering that out is critical. It's something that we do here. We have a handful of agents that we use, but we've tested them relentlessly and something I've heard everyone here say, human in the loop. I don't see, at least with current technology, a way that you can remove the human. Somebody has to be
accountable. Somebody, even if you engineer the perfect agent, has to evaluate it for drift, for hallucinations, whatever it may be. And that's how we've designed our product. And it's not just about us, when we look at other software that is AI enabled or AI native, the first thing we ask for is what's the embedded control? Where's the human in the loop? How do you know it won't drift? What's the checks and balances, the guardrails, whatever word you want to use, that's absolutely critical out of the gates that that has to be built into the technology.

Tom (13:47):
Yeah. Patrick, I'm going to stay with you and shift it slightly here because only between 15 and 20% of CFOs have actually fully embedded AI into their operations, though that number is up from less than 1% last year. So as a CFO yourself, why is there a gap between ambition and implementation? Why is that gap still so wide? And how much of that is genuinely a legacy systems problem rather than say a strategy one?

Patrick (14:15):
It's a little bit of both. And I'm not speaking personally, but I have a pretty broad network of CFOs and CIOs that I speak to and was brought up earlier, but you can't release agents into your environment until you feel good about the integrity of data, right? If you release agents into a pool of data that's inaccurate or disparate or is reflected differently in various different systems, that agent's going to return several different answers that don't match. So part of it is legacy systems. You have to be in an environment that you know your data is a single source of truth, you know that it's accurate, and then you can release an agent into it. So that's part of it. I would say another part of it is everything that we've been speaking to earlier here, and I think it's a good thing. That's why we went from 1% to 20%.

(15:08):
I bet that number next year is well over 50%. It's going to move at a hyperbolic rate. But we have an obligation, especially as a public company, that anything you use that impacts financial information, you have to test that on a parallel basis to multiple periods, if not quarters. So you have to get your internal auditors aligned to that. You have to get your external auditors aligned to that. You have to design and change controls around that. You have to reevaluate your processes. Maybe historically you said, "I would never do something this way because I'd have to have human beings working 24 hours a day and that's completely inefficient and ineffective." Well, now that's back on the table, right? Oh, I can have an agent run through the night and do this maybe a different way. So there's so many different factors you have to evaluate with this technology and be careful about it and be disciplined about it.

(15:59):
I think that's why the rate of adoption is twofold. It's one, legacy systems, but it's two, in our world, you have to test it for a while and get all the constituents aligned before you go live with something or release it into a live environment.

Tom (16:15):
Olia, let's stick with this. And in fact, I'm going to continue with my plane analogy from the top of the show. Is the task ahead for CFOs much like building the plane while flying it, replacing siloed spreadsheets and manual reconciliations with a single unifying data architecture, if you will, all while the business keeps running? And what does this actually look like in practise? Because you can't do this without disrupting the day-to-day.

Olya (16:39):
Welcome to my life. This is really what we have been working on, and I do believe that it takes time with proper investment and talent and reasonable timelines. It's actually very possible to do it without the interruption. But when you talk about the unified data architecture, it's probably the hardest and the most time-consuming part. We started a project three years ago where we had to go through the full data transformation process. We changed our chart of accounts, we re-implemented our ERP, we rebuilt all of the integrations, moved our data warehouse to the cloud, and all of this have happened in the background. People didn't even know what was happening other than the newsletters coming their way. And then we had to go through the shift where you go from the old to the new, and really all that work had to be unnoticed. We put so much time and effort into it, but with the proper change management, with the proper communication plan, and with running some systems in parallel for testing, just like Patrick said, it's very, very possible.

(17:49):
Avoiding disruption just means having a good execution change management plan.

Tom (17:53):
Gina, traditional software runs on if then logic. Agentic AI runs on probabilistic reasoning. If an autonomous agent triggers a hedge that costs the business millions or denies a credit line based on skewed data, the question of who's responsible becomes a genuine fiduciary nightmare. Privacy and ethical risk are major concerns, largely driven by fear of the reputational cost of a rogue AI decision. Practically speaking, how do CFOs explain to a board or a regulator exactly why an autonomous agent made a specific call and where does the responsibility lie? Is it the developer, the organisation or the human in the loop?

Gina (18:36):
Yeah, this is an interesting question that's come up several times and I really appreciate it. I think the 30 odd years I've been advising CFOs, this has really become the pressure point, but really there's no excuse. The accountability has to stick with the humans. As Harry S. Truman said, one of our dear presidents, the buck stops here. You're the CFO and you're signing the name to the financial statements. That's it. You could use all the excuses you want. If you don't manage and/or monitor or have governance in place to ensure that doesn't happen, where is that check and balance? Where is that accountability? The system doesn't make a decision on its own. Somebody allowed it to make a decision. So it was programmed, it was allowed to, it was monitored at the right level. So for me, it is the CFO's and the explanation is sometimes you have to take responsibility for whether you and your organisation didn't address it, didn't identify it first, and it became an issue.

Tom (19:32):
Patrick or Olia, would you like to jump in here and add any insight?

Patrick (19:37):
I would just say Gina's spot on. There is no way a CFO can walk into an audit committee meeting or a board meeting after an agent just did something inappropriate or made a mistake and not be accountable for it. It's our signature. It's our responsibility. You can't blame the CIO. You can't blame a software vendor. It is our responsibility to validate the accuracy of everything that is generated either in a spreadsheet, manually, out of a system by an agent. It's all the same mentality. It's just a different technology and we own that. So that's why, going back to your previous question, it's so critical to never even put yourself in that position, to have an agent that could go that rogue and do something that
materially adverse to your business. That's why you need the human in the loop to prevent it from ever getting to that point.

(20:22):
That's why you need to quote engineer that out before that agent's even capable of doing that because you don't want to find yourself in that position. Even if you are accountable to it, you don't want to be in that position. Leave it at that. Olia. You

Olya (20:33):
Heard me say no human owner, no go live. I absolutely agree that human responsibility in the work of AI should always be there. And if you look at all of the emerging legislation, that's the requirement. So I think paying attention to all the legislation that is coming out from Europe, from China, here in US, it's a non-negotiable.

Tom (20:57):
Well, with the importance of humans in the loop when it comes to those responsibilities and decisions, I want to look about the people part of this process a bit more. As AI takes over level one tasks, there's a fear that entry level finance roles will be hollowed out, but it's not just finance. There's lots of different industries which are kind of experiencing this same rhetoric and this same conversation. Junior analysts used to learn the language of business by doing manual reconciliations and data entry. If they're not doing the grunt work anymore, how do they build the intuition needed to becoming senior leaders one day? Patrick, let's start with you.

Patrick (21:35):
I don't think agents are going to completely replace level one or entry level individuals, and you're absolutely right. That apprenticeship model right out of college when you start your career in accounting or finance is critical to build to where you are today. I remember being a first year associate and you're footing things, you're reviewing reconciliations, and that's how you learn the basics, the bones of accounting and finance. So that I don't think will ever go away because you're still going to need that knowledge. You're going to need that knowledge to be able to review the work of agents and know that they're performing their tasks accurately. Now, what I could see is maybe the pyramid, which is maybe like this right now with the CFO at the top and a large, large team underneath. I think that pyramid will become much more narrow at the bottom, at the base, because maybe the days of needing thousands and thousands of staff accountants, they go away, but that's not necessarily a bad thing.

(22:31):
For one, there's a CPA shortage, at least in the United States. There's less and less people joining the profession as being harder and harder to recruit. But I also think maybe the talent at the base of the pyramid at org structure will be different too. The days of, "Oh, I worked in the big four and I have a CPA and I'm a technical accountant, hire me." I think those days are already starting to become behind us. I think CPAs are going to have to be data oriented. I think they're going to have to be more tech savvy. I think they're going to have to understand AI. I think it's an entire skillset that they're going to have to develop, which will make them more than just accountants within an organisation. So like most technologies, just 40 years ago, 45 years ago, Microsoft Excel was invented and everybody said, "Oh, we're never going to need accountants anymore.

(23:19):
The computer's going to do all the work." Here we are. 20 years after that, companies like BlackLine came about, "Oh, we're automating Excel. It's all going to be in the cloud. We don't even have to put stuff in Excel anymore. All the accountants are going to go away." No, our skillset changes, our technical proficiency changes, and we just pivot from a career standpoint. And I think that's what's going to happen with AI.

Tom (23:41):
Yeah. Gina, I want to look a little bit at the mentorship because Patrick, as he described, there's a shift from a pyramid and many juniors, few seniors to this middle heavy diamond. Is that structurally sustainable or are we just deferring the mentorship problem by say a decade?

Gina (23:58):
I think the pyramid is definitely shifting to similar to what Patrick said, but it is about real world experience. So whether you have 20 new consultants starting or you have 10, giving them the opportunity to roll up their sleeves, do the work, work alongside of the tools that are there, and I think of AI and the agents as tools, just as Patrick said about Excel. Excel is a great tool that came about, but it still needs the human interaction and using judgement , teaching communication skills, how to get people on board, really understanding what those key metrics are. You need to be able to teach them. And I think the mentorship programme, especially in this environment, is going to be critical. So I think that the pyramid is maybe not a pyramid anymore, but it is certainly not completely disappeared.

Tom (24:46):
Yeah. Thank you very much, Gina. Olia, as I said at the start of the show, the technical half-life of a skill who shrunk to just two and a half years, how do you budget for a workforce that needs re-skilling on a near constant loop?

Olya (25:01):
It's a great question. I would continue to budget for experienced subject matter expert and then compliment them with the AI natives. I've seen the magic that can happen when you peer up an experienced practitioner with the Wizkid who knows how to use the technology and really continue to invest in training and AI literacy because again, A, it's a requirement, but B, you want people to understand the tools that they're using so they can have the ideas of how you're going to be using that going forward. I would continue to hire for attitude and curiosity. My whole philosophy is work yourself out of the job so you're ready for what comes next. And two and a half years is actually a really good cycle to be able to do that. AI allows us to move faster and make that cycle even shorter, but those individuals with natural curiosity and learners would be the ones who would be the most successful.

Tom (25:58):
Well, look, we are going to come to some closing quick fire questions. So if we can come with some brief responses, I know they're big questions, but let's have a crack. What is the one roadblock, technical, regulatory or human that could stall the shift to a real-time enterprise and how do we get past it? Patrick, you first.

Patrick (26:18):
I would say it's to get to a real-time enterprise, you need absolute data integrity. And that sounds easier than it actually is, especially the bigger you become as an organisation. And I know you want this to be
pithy, but when you have multiple ERPs, when you have multiple data warehouses, that is extremely challenging. And for AI to work on a mass scale, agents to work effectively and accurately on a mass scale, that's a major undertaking for a lot of large multinational companies. Julia.

Olya (26:55):
I second that. Data readiness and governance are the two big items in my opinion. And while you focus on building that data infrastructure, I think you can lean in onto the partners that you're currently working with. We are partnering with BlackLine right now because they have done all that work and they have invested in their technology, so now we can reap the benefits. And with the work they have done going through the ISO 42001 certification or going through the SOC one work, we can rely on that rather than doing it ourselves.

Patrick (27:28):
Gina?

Gina (27:29):
I think the major issue is change management. Adoption is really about the people that are using the tools are going to use. Right now, what we see and what's slowing it down is the fact that people are fearful. They're fearful for their jobs, they're fear for how they're going to be working with it, their fear that the tools that we're putting in place are going to replace them. So if you can help manage that fear and really manage the change management, not just within the finance organisation, but really outside the finance organisation, and the CFO has the opportunity to lead the change, to be the driving force and say, look to the rest of the organisation, these tools work, this is how we can do it. This is how we can make faster decisions, better decisions with more data quickly and enable to be more enlightened as to what they can make decisions on.

Tom (28:17):
In five years time, will the average finance professional feel liberated by their AI colleagues and look back on this time like we did 30 years ago and say, wow, we're in the dark ages in comparison to now? Will there be this nostalgia for a spreadsheet and a chat with a colleague? Patrick.

Patrick (28:36):
So I think just like any other technological wave, the professionals that lean into this that aren't scared of it, that embrace it, five years from now, they're going to look back at how they used to do their jobs and feel relieved. I think they're going to have a much better employment experience. They're going to be more analytical. It's going to be a better career, and I feel firmly about that. And yes, there's going to be professionals just like in the early '80s that said, no, I don't want a desktop computer. I'm going to continue to do my job with a 10 key calculator and green ledger paper. And those people change professions by the mid '80s. So you either adopt it, you lean into it, you embrace it, and I feel very confident that those that do that will have a great experience. Julia.

Olya (29:20):
I really think that our jobs will definitely change, but I don't think we know how yet. If you think about the positions and the roles we have right now that are built around the requirements that are out there for us, a lot of companies or a lot of governing bodies, legislation and AICPA right now are deciding what those new requirements are going to be with the growth of AI and things are changing so rapidly. So five
years is a very long time. I think we are going to start seeing those changes this fall when all those governing bodies are going to start coming with the new requirements for the work we are doing. But at the end of the day, everything will continue Continue to evolve like it has been for the past many, many years.

Patrick (30:03):
Gina?

Gina (30:04):
I think it's just another tool in the toolbox. It's going to be integrated. It's going to be part of it. I love what Patrick said. I did work on 18 column green pieces of paper and then got a computer and there's no way I want to go back in time to green pieces of paper as I follow cash around the globe. I'd rather do that in an Excel file and I'd rather do it with another tool that could do it even faster, better, and have more accuracy. And it really is just another tool in the toolbox. It's just about that learning process. It's that education. It's that comfort level that people need to get in order to look back and say, oh wow, I can't believe we live without this.

Tom (30:39):
Well, our final question, and it's a big one, but what is your one hope for where we will be in five years time? Patrick.

Patrick (30:49):
Boy, that's got to be a succinct answer. My one hope. I would like to see almost an entire elimination of the rope tests that accounting and finance professionals currently endure on a regular basis. It's gotten better over the last several decades, but we're not there yet. There's still a lot of stitching information together. Even though we work diligently to eliminate the need for Excel, I don't know of an accounting and finance org in the world that doesn't use Excel. It's just the nature of the beast. I would like to see a world where a lot of the work that you have to do just to get to an answer before you can evaluate it, before you can analyse it, that most of that work is automated so that you as a professional spend more time bringing value to the business, bringing analysis to the business, bringing advice than just compiling that information.

(31:41):
And while we are exponentially ahead of where we were just a couple decades ago, we're not there yet. And I think AI could be that missing link. Olya.

Olya (31:51):
My hope is that responsible AI helps us create capacity. Capacity to live our lives, capacity to take vacation, capacity to spend some extra time with the families, not just focusing on the work. I think it's easier to say where I don't want to be and where I don't want to be is I don't want to see increased workloads or shrinking timelines or growth and responsibilities because there are a hundred agents constantly doing the work that you have to review. I think we have to be very, very careful not to let AI take over our lives in a sense of having all that work that human in the lobe becomes the bottleneck for it.

Patrick (32:32):
Gina.

Gina (32:33):
I would agree. I think the hope challenge is for AI to just no longer be treated as something of a threat, but just really be incorporated. And that it is really a trusted tool that's understood and utilised to the best of its ability. It's a real opportunity, as I said before, for CFOs to lead this. My hope is that the technology provides an improvement in what we produce that makes the organisations

Tom (33:01):
Better. Well, a big thank you to my guests, Patrick Villanova.

Patrick (33:05):
Thank you. I appreciate it, Tom. It was a pleasure meeting everyone. This has been a great experience.

Tom (33:10):
Olia Brasi.

Olya (33:11):
Thank you. It was a pleasure to be a part of this conversation. Thank you for inviting me.

Tom (33:15):
And Gina Goodsight.

Gina (33:16):
Thank you so much. It's a great conversation. Appreciate the time.

Tom (33:23):
Well, as we close this chapter, it's clear that the CFO's job description has been rewritten. This was never really a conversation about software. It was about whether an organisation can hold onto the rigour of the past while reaching for the speed of the future. We've talked about human interactions not being replaced, but about the unglamorous but essential work of paying down workflow debt and about a talent pipeline that has to be rebuilt from the ground up. If there's one thread running through everything Patrick, Olia and Gina have shared today, it's this. In the age of the autonomous enterprise, the CFO's greatest asset has never been just their capital. It's having a team of humans ready to act on insight before it becomes history. I'm Tom Parker and this has been The Next Five Podcast. Thanks for listening.