Finance specialist discusses the “remarkable” impact of AI on finance, while urging caution from CFOs

Carl Seidman is a finance specialist who has spent his career in management consulting and FP&A advisory work. He’s the founder of Seidman Financial, where he advises and educates businesses of all sizes in 24 industries.

We asked Carl about his experience integrating AI into finance advisory work across a multitude of organizations. Here’s what he told us.

Every finance leader’s responsibility

I’m Carl Seidman. I’m a finance specialist with a core focus in strategic FP&A, turnarounds and restructuring, and FP&A team development. I have spent the entirety of my career in management consulting and FP&A advisory work. I’ve worked with companies of all sizes across more than two dozen different industries. They’ve gone through every sort of circumstance — from startup and capital raises to rapid growth and scale, maturity, financial and operational restructuring, bankruptcy, and exit. And everything in between.

I am the founder and principal of Seidman Financial. We work with companies of all sizes and across more than two dozen different industries. What began nearly two decades ago as a financial advisory business serving companies that otherwise didn’t have access to this level of midsized-company finance expertise has evolved into a premier provider of financial education and training.

Where I find myself today is at the intersection of finance and AI. It’s understanding not merely how to prompt and use the technology, but how to really embrace its capabilities in real-life applications.

I’m not alone. It’s impossible to be in a finance and accounting environment today and not be actively thinking about what AI means for the future of our work and companies. As someone who has extensive connections to many FP&A software companies, I have firsthand visibility into what these software companies are building and how they’re rapidly integrating AI into the core suite of solutions. Platforms that used to focus on more effectively managing finance and data are now looking at ways to augment their offerings for this new AI reality.

On the people front, many finance and accounting professionals don’t know what they don’t know when it comes to AI. For years it was a non-issue, yet today, it’s on everyone’s minds. And because of that lack of familiarity, as well as the extraordinarily fast pace of change, so many feel like they are already behind.

The fact of the matter is that AI is never going to go away. Which means we have a responsibility to understand how it plays into our work. But at this stage, it’s anyone’s guess what finance and accounting powered by AI looks like a few years from now. The best we can do is to continue to learn, stay informed, and connect with the people who are a few steps ahead of where we might be.

How Claude changes a financial leader’s focus

We have now embraced AI tools — most notably, Claude — for many of our own workflows.

Since the introduction of Claude in Excel, Claude in PowerPoint, Claude Cowork, and Claude Code, we are now able to automate many human responsibilities — and at lightning-fast speed. The output is as good as, if not better than, what a human could produce.

For example, we regularly use Claude in Excel. I’m a Microsoft MVP (Most Valuable Professional) in Excel, and I have extensive financial modeling and analysis capabilities. But even for me, it’s more efficient to use AI. What otherwise may take me a few hours of intensive focus can be done to 90% of the quality in 10 to 40 minutes.

And it’s not that I’m forgoing my involvement in the build — it’s almost as though I’m the coach and I have a really smart staff person sitting next to me who can do everything I’m asking super quickly and super well. This allows me to save my mental energy for higher-value tasks, as well as to focus more on what the models are telling me.

In the end, that’s what business partners and clients want from us. They want us to be supporting them in their decision-making, not deep in the weeds building financial models.

How Claude Cowork can power thought leadership

Here’s another example. We use skills embedded within Claude Cowork to enhance much of our thought leadership.

Nothing is created from scratch within Claude, but we feed it the rich content and provide it the context to make it even more robust.

So, again, we get to use the technology to make our work better, but it does not replace us.

The remarkable output of Claude

What’s perhaps most remarkable is just how good Claude is at doing finance and accounting work. Years ago, it seemed like AI was going to fall short because it wasn’t performing our duties at even a 50% competence level. But since then, the precision and capability have astounded me.

Not only can it work within the model and run scenarios with high reliability, but it can also converse with us. If we turn on a voice-to-text app, we can literally have a conversation with the tool. We can tell it what we want to consider, ask questions where we’re not sure of the answer, and the tool will respond to us in a thoughtful way.

Here’s a recent example. I was modeling the operations of a furniture manufacturing company. In 40 minutes, Claude helped me build a 17-tab financial model that otherwise would’ve taken me hours, if not days, to build myself. It wasn’t exactly the way I would have built it — it looked like a junior staff member had made it — but the model was absolutely sufficient for the client. When I incorporated the scenario management engine, it took the form of an interactive dashboard.

What that means is that I can meet with a client and facilitate a discussion, or the client can manage it themselves. And we get to do this live instead of having a meeting, going back to the drawing board, restructuring the model, and discussing it over again. All of this can happen in a matter of minutes.

The greatest disappointment with AI

Now, with all that said, perhaps my greatest disappointment is that the AI still acts like it’s a junior employee.

As I said, that is sufficient for many work instances. But where we really might want AI to be far more robust, it can’t handle it — it’s not the advanced dynamic financial modeler that we’d like it to be. It may get there eventually, but it’s not there today.

So, for basic model builds and analysis, AI is a great tool. But for more complex modeling and analysis, we are still sticking to our own capabilities.

Why CFOs must be cautious with AI

My advice? Always be learning and always be growing. The space is moving extremely fast.

I also caution CFOs against buying into the buzz and marketing. It’s not worth investing huge sums of money into systems that are still immature. The ROI on the finance side is still not where many AI tech companies claim it is.

So, be extremely clear on what it is you’re trying to accomplish and ask yourself whether AI is truly the solution to help you get there, now or in the coming years.

Follow along

You can follow Carl Seidman’s work on LinkedIn. And check out Seidman Financial.

More expert interviews to come on The CFO Club!