How Generative AI Helps FP&A Pros Replace Legacy Systems and Boost Accuracy with Natalia Toronyi

Welcome to Future Finance. I'm one of your hosts, Glen Hopper, and our other host, Paul Barnhurst, is with me here today, and our guest today is Natalia Toroni. She's an executive with almost 20 years of experience leading financial transformation across Fortune 500 global industrial manufacturing companies. Natalia is known for her sharp problem-solving skills, curiosity, and transformation mindset, especially when it comes to driving impact through AI, automation, and data. She's led major restructurings, M&A, and operational turnarounds, all while staying grounded in integrity and leading by example. She's also a passionate mentor and a strong advocate for people first leadership, always focused on building trust, empowering teams, and creating cultures where innovation and inclusion thrive. Natalia, welcome to the show. Thank you for inviting and the nice introduction. Let's start here. We'd love you for you to share a little bit of your journey. I know you live in Ukraine, Budapest, now in the US, you've worked for several different companies. Just tell us a little bit about your journey and background and how you ended up where you're at today. I was born in Ukraine in a small town called Oshweroth. It's in the Western Ukraine, and I grew up during the very interesting time when Soviet Union collapsed. It was a very, the time of chaos. I lived through hyperinflation with daily currency revaluation. People worked for six, nine months without getting paid until the majority of the companies just completely collapsed and shut down, and everybody had to figure out how to survive and how to live. When I'm looking back, I still don't know how my parents managed to put the foot on our table every day. For us, it was just normal. We were kids with and paid too much attention to it. But now, when I think back, it always comes to my mind, that was the great lesson to learn no matter what's happening around you. Giving up is not an option because you have to survive. In my early 20s, I moved to Hungary. I graduated from Ujshvar National University with major in accounting and audit. When I moved to Hungary, I realized I actually have to start from the very beginning from the scratch, because not only my degrees from Ukraine and I need to now revive, revive, revive it. I need to start the language from the very beginning, which is Hungarian is one of the most complex languages in the world. So, I started to learn the language from the cartoons, which at that time my daughter was watching. The second thing I learned in very small details was the law about accounting and audit, because I was ambitious and I wanted to continue my career or restart my career. I had to deep dive into how can I do that. So, I revaluated. I had my diploma, revaluated and I kicked off my finance career in Hungary in a small CPA office, where I was the best employee, because I could not participate in any other discussion, then processing the numbers, but you know, that gave me an opportunity to learn the language and improve it every day. So, shortly after I joined my first corporate company, Alkoa, where you know, with help with excellent teachers and mentors, I learned, I first was exposed to USGAP. I was given a better visibility of what it meant to operate within global processes, what sucks compliance means, and how to contribute to performance in of the global corporate company. I think that company Alkoa was had the best and cleanest processes I've seen during my 20 years, with a huge focus on SOD segregation of duties and sucks compliance. So, that was my base, which gave me a very strong step into finance. After this, I joined multiple companies, changing, growing in my career through BPE, Eton and Signal. I led global finance teams, delivered numerous transformation projects, you know, implementing different automations now, AI. I helped to build finance operations from this scratch, centralizing multiple departments into one, guiding finance departments through cyber incidents and fraud recoveries. And always in my work, I always champion innovation automation, data analytics, and better processes. So, that's shortly my story. That's amazing. And I think that the adaptability that you had to have in moving to a new country, learning a new language, learning accounting in that new language had to be just so much to work through. But it seems like what the takeaway from that is, is that you learn to be resilient, you learn to be adaptive. And that's, I mean, things that you would apply for the rest of your career. So, you know, putting in the hard work early and going through that, it seems like it's served you well through the rest of your career. Well, going through all that, I mean, you know, future finance and Paul knows every time I get up, the first question I ask is it going to be about AI. So, as soon as you mentioned AI, I kind of locked on to that. And I'm wondering with your journey through accounting and finance and all the transformations you've done, how are you seeing the latest generative AI and technology reshape the way that you work today? And will you work in the way that the profession is changing and will change? I think it's completely reshaping industries and, you know, industry and all the departments, not only finance. A lot of times people are asking questions, you know, which department will be impacted, which department will not be impacted. I can see even looking into finance organization, every department will be impacted. And we already saw a lot of automation in complete order to cash process, be procured to pay process, GL cost accounting. I can see a lot of departments transitioning and merging, where now they will be more complex and more connected. So, AI is bringing opportunities, you know, to different analytics, it gives us visibility. It changes how we forecast. It's providing now a different business insights, which before we didn't have. I think it's shifting the departments and changing the entire departments and how we do the process transactions. Yeah, absolutely. I agree. You mentioned, you know, consolidation, really maybe some complexity. There's just so much going on. And I'm curious, I know you've worked in Treasury. What's your thoughts in the Treasury space as far as AI? Are you seeing, or do you think we'll see a lot of benefit around cash flow forecasting, you know, working capital optimization, FX risk? Would love to kind of get your thoughts in that space of maybe, you know, what you've seen and what you think we'll see there? You know, I mentioned a couple of minutes ago that departments will start to transform and change the way how we are pulling the information and analyzing that information. You mentioned cash flow forecasting, which historically was a Treasury function and working capital, which is ARAP, inventory management. I think those two processes will now be more blended and more merged, because now we will not use only information from the bank statement to focus. Cash, but we will deep into our ERP systems, gaining visibility of, you know, customer behaviors, vendors, terms, and how we schedule inventory demand. All those activities will be transformed and changed on a more precise data point. We can now put, incorporate macroeconomic indicators. We can build our models more precisely using all that information. Yeah, and you know, when you talk about bringing in macroeconomic data and sort of pulling in data other than whatever drivers you have internal to the company, that makes me think back to, I guess we'll call it classical AI, the more machine learning, what's been around for years? And a lot of people just today are just talking about generative AI, and of course that's taken the world by storm, and it's made classical AI more approachable and accessible, because you don't have to get through it through Python and coding and all the barriers to entry we used to have. And as I travel around and talk to different groups, an area that is, I was going to say near and dear. I don't know if that if audit is near and dear to anyone, but it's going to say when did audit near and dear? I usually run from the auditors. I say that out loud. So near and dear was the wrong choice of words. Audit is something that anyone in the public company space certainly is very familiar with, and it's an area that is very time intensive and very, you know, a lot of the work that we're doing in an audit is, or for the, on the auditor side as well, is very, it could be repetitive and it's complex. And I get asked all the time, is AI going to help with this? And the flat answer is, AI is going to help with everything, but I don't know, you know, exactly when. But I'm wondering from your perspective, how do you see AI helping manage the audit? process. Do you have tools that you're using today? And that could be generative AI or even, you know, classical machine learning. I think probably one of the first starting points, which I would implement, if I would start utilizing, you know, AI in multiple finance processes would be internal audits. And I would start with creating well-defined written processes and controls. Probably the first option, which comes to my mind, would be Gen AI through OpenAI, creating the GPT, uploading the clear defined processes SOPs into that AI agent or assistant, and starting to create the instructions for all the, you know, departments. And based on that, also utilizing the same document created and and cleaned for internal audits or for for pulling information for the audits. I can't even say that it would be annual audit now because before we would create some, you know, processes where we would use random samples selected for auditing. Now this quick, fast tools, they give us an opportunity to test not only samples or not only 30 or 50 samples a month, but we can test, you know, 30% of the samples or 50% of the information to look for deviations, to look for exceptions, and only pull those which don't align with those clean, defined processes and SOPs. So I didn't see and I don't, I can't mention any AI tool which is helping with the audits right now, but that would probably my main focus because from there you can build all the processes correctly without, you know, that's your base. Yeah, it goes hand-in-hand with controls and it's funny as I talk to people, I don't know what it is, and Paul's probably heard me say this before, but I'd go and I'd talk to a room full of CFOs, controllers, auditors, you know, just all across finance and accounting and it was always an auditor who came up to me and would say, "Hey, I could never replace my job." And I always think, is there anything more rule-based than audit? You are perfectly set up to be replaced by the bots. And if Paul, if you can let me get up on my soapbox here, what I love that you said, go ahead, get on your soapbox plan, it's been a while. Natalia, I loved what you said about documenting processes. So I got my career start in the military, and in the military you had to have standard operating procedure for everything you did, and I have every role I've had. I'm documenting what I do, and then when I had direct reports, and all the way down through my departments, we're going to document everything. And now, you know, seeing where we are with AI, I feel like, "Well, that was a really prescient approach. If I do pat my soap on the back here, because the same SOPs that we would use to train new employees, well, guess what? Those are going to be used to train agents that are going to be doing this job in the future. So if you haven't put your SOPs together, this is my soapbox, Paul. If you haven't put SOPs for everything you do that now, because in the very near future, we're going to be using those SOPs, not to train new employees, but to train the bots who are going to take our jobs, I guess. Yep. So if your SOPs are not correct, the AI will generate the full cows everywhere. We've talked quite a bit about AI, a little bit of treasury audit, different areas in finance. But I think something a lot of leaders struggle with, and I've had them ask me, "How do you distinguish between what's real and what's hype?" I think I saw the other day, in the last two, two and a half years, whatever it's been now, two and a half since version 3.5, a Chichi Pika amount. There's been 75,000 AI native companies start. Now, not all those are in finance, but just think how much that is. I think people are just overwhelmed. So how do you look at that and evaluate technology? What advice or thoughts would you give for somebody who just looks at it all and goes, "I don't know what's real, I don't know where to start, they're just overwhelmed?" It is. It's a great question. It's very easy to get lost with all the solution offered right now. Pretty much every second company is offering some AI solutions, which in a lot of cases it's not even AI, it's machine learning, which was there before. So we need to make sure that we distinguish well between marketing, tool and the real AI and machine learning involved. In the processes, but I always come back to three things. Does it solve the problem, the real problem I currently face? Is it scalable and can it be measured? I look for solutions that, is there a safe time? Bring some efficiencies, reduce some risks or help in decision making in different processes. Adds the complexity to already existing processes and it requires additional multiple steps and we don't have clear return on investment. It's an all goal. Another test would be probably how that particular solution plays with our existing systems and how also very important how it comply with our security requirements and SOX compliance requirements. In the past, those solutions often require the replacing or upgrading legacy systems, building very complex integration, cleaning data, migrating data, which would lead to high infrastructure costs. I'm not an expert, I'm not a technology person, but based on what my research is, it seems like with AI, it's changing because now AI, especially if we use different Gen/AI, RPA, it reduces the need of full system implementation. So it brings potential solutions to the companies which don't want to invest into ERP re-implementation or bringing new ERPs and it still can bring efficiencies into the processes. You need to know again, going back to clean processes and SOPs and defined rules, we need to know what we expect from that tool to solve because if we just think, "Oh, I'll bring AI and it will solve all my problems," it probably will create more chaos than it solves problems. Perfect, I love that answer. All right, well, this flew by fast. So we've covered everything we set out to and I think that some great insights and Natalia really, really appreciate you coming on the show. Thank you for inviting. It was very interesting to share my perspective. Appreciate it. Thank you again for joining us. It was a lot of fun and we look forward to having your audience get the opportunity to listen to this. I'm sure they'll love it as much as we did, so thank you Natalia, really appreciate it. Thank you.