These days, it seems like everything is overhyped. From the latest movie release, newest iPhone model or even the weather, every media outlet and social media post wants a click, creating a loop of endless hype cycles.
Especially in business, that can make it difficult to decipher which piece of technology or innovation is going to be the ‘next internet’ and which is going to fizzle out like NFT’s. Generative AI for treasury management has been the latest buzz in finance, but are there real business use cases for integration into your finance tech stack, or is it all smoke and mirrors?
In our view, there is absolutely value in AI, and there are countless real world uses for the technology that savvy finance teams are already taking advantage of. In fact, according to a recent survey from KPMG, 65% of finance executives say they’re using AI in some capacity to help them lower employee workloads and to achieve better data accuracy, reliability and predictability.
So, full steam ahead then right? Well, not quite. As much as AI use is exploding in the finance space right now, not everyone’s on board. A recent report by the AI Accelerator Institute found that almost 50% of users of AI technology don’t actually trust it. Not ideal.
The value of AI is definitely there, but with trust issues and uncertainty around use cases, more work needs to be done in demonstrating its practical applications in clearer terms. So let’s do just that.
What Does Generative AI Mean for Finance Teams?
There’s a healthy amount of skepticism among treasurers and CFOs when it comes to generative AI. We’re seeing many taking tentative steps to integrate the tech into their teams workflow, but at the same time feeling unsure on the real benefits of the technology.
A big part of that is a marketing problem. AI isn’t magic, and despite what the LinkedIn influencers would have you believe, it’s actually not brand new. Simply put, generative AI is data analytics technology. It has the capability to sift through vast amounts of data to identify trends, spot anomalies or find specific pieces of information in a vast sea of numbers.
For finance teams who deal with sometimes millions of line items, that makes AI perfectly positioned to offer immense value, saving your employees time and minimizing errors.
Broadly speaking, one of the overarching key benefits for treasurers and CFO’s is in AI’s ability to automate cash and finance reporting. Rather than tasking an employee with building a specific financial report that needs to be compiledf manually and could take hours or days to create, generative AI can build a report in seconds, compiled off data that is real time and accurate.
It puts finance leaders closer to the numbers than ever before, meaning less time spent on getting cash figures or financial numbers to report to other executives. This time saving allows treasurers and CFOs to add more value with strategic advice and analysis, making their role even more valuable to the company.
But let’s get more specific and take a look at some common integrations for various finance functions.
Check out our recent episode of Fintech Corner, as Joseph Drambarean, Brett Turner, and Jeff Macke discuss everything you’ve wanted to ask about AI; What’s so special with this new wave of AI and are finance teams being impacted? How do finance teams change the way they approach their jobs with AI in the picture? How to tell if it’s just a shiny new object or a real, here-to-stay evolution? And last but not least, how can a finance professional use it at work without being fired for compromising their data?
Financial planning and analysis typically requires a significant amount of manual work. Before any projections or models can be built, financial data needs to be collated, organized and reviewed for accuracy. This can take a substantial amount of time. AI can assist in creating an accurate baseline scenario, with next to no manual input.
Data feeds can be set up through the use of cloud native software, open banking APIs and accounting integrations to collate financial information in real time that is 100% accurate, and AI can then be used to organize this data and find anomalies.
This moves the FP&A team away from building the foundation of financial data through manual data entry, freeing them up to spend more time on high value tasks like strategy and analysis.
From there, there’s no longer any need to build models from scratch, and to rebuild them for a range of different scenarios. Using natural language prompts, software add-ons like Trovata AI can be used to build a model using normal phrasing, rather than via complex spreadsheet formulas.
Instead of needing an MBA in spreadsheets, your team members can simply ask something like “Show me these projections with a 5% increase in average sale price and a reduction in average finance interest payable of 1 percentage point.”
Where FP&A is all about financial models that encapsulate broad corporate financial data, transactional finance is about the finer details. For large organizations, connecting payments, reconciling accounts and managing accounts payable and receivable can be a mammoth undertaking. Particularly now, when this information needs to be gathered from so many different sources.
Here, AI’s capabilities in sifting through huge datasets in an instant is a valuable tool. The technology can be used to recognise and match invoices to payments, automatically reconcile accounts and track overdue accounts.
It can highlight trends in payment timescales, monitor transaction fees and build reports with automation and natural language prompts. And if you need to find specific transactions, you can task the AI with combing through every line item to find them, rather than your employee having to do it manually.
One of generative AI’s strongest features is its ability to identify trends within a dataset. While trawling through multiple quarters of financial statements to pull out interesting changes in financial metrics like gross margin, operating cash flow or cost of goods sold might take a person days, AI can do it in seconds.
This can help create an excellent starting point for investor communications, highlighting areas of outperformance and areas for improvement. Many AI integrations can even put some basic commentary on why these metrics have changed. For example, an overall jump in revenue could be attributed to an increase in sales in a specific region or product line.
Your comms team can then review and adjust this messaging, but it gives a solid foundation with very little work.
All of these use cases are summed up perfectly by Aurelia Sirbu, CFO at Orbus:
Integrating Generative AI for Treasury Management the Right Way
That all sounds great, but what’s the catch? First, we need to talk about trust. As mentioned at the outset, the trust factor when it comes to AI is a big issue. Some executives are concerned about the accuracy of the outputs from generative AI, and others have some reservations about the security risk of sharing data.
On the first point, it’s key to understand that specialist software tools, like Trovata AI, don’t have the same issues as well publicized large language models (LLMs) like ChatGPT. Those have been trained on an impossibly large dataset, which includes sources as broad as affiliate blog posts, Reddit comments and Tweets.
With that comes a huge amount of areas it can offer answers on, but also a high likelihood of conflicting information in the training model.
But point the technology at a narrow dataset like the financials of a single company, and many of those issues go away. The model isn’t trying to offer you the best answer from a sea of different answers, but is instead finding you the one actual answer in the dataset.
As for the security aspect, your financial data remains entirely secure with Trovata AI. The LLM integration is used to transpose your natural language prompt into a specific data query, without the actual financial information ever being passed through it. Think of it as a translator, with the answer to your question sent directly to you.
Even knowing all this, finance executives should expect it to take some time and money in the short term to integrate AI into their existing workflow. Once users get a taste of the capabilities of AI, it can be tempting to want to overhaul the entire process to maximize AI integration. While this is worthwhile as a long term project, it’s important to go slowly to minimize disruptions and to ensure everyone on the team is comfortable with the new ways of working.
It’s likely to be time and money well spent and many are taking the plunge, with a Gartner survey from 2022 showing that 80% of surveyed CFOs expected to increase AI related spending over the next two years.
Implications for Finance Talent
So with generative AI, this means you can operate your entire finance department, without a single person on the payroll….well, not exactly. But this is one of the big misconceptions around generative AI, and another favorite of the LinkedIn and Twitter/X AI bros.
The reality is that AI is a tool, and one that will continue to need real people to use. Much like the internet in general, new technology won’t lead to overnight mass layoffs and the end of work, but it does create the ability for smaller teams to get more done.
That means being able to hire slower, and giving less technical members of the team the ability to do tasks which they wouldn’t previously have been able to do. Finance teams can run leaner and move faster, and with less time spent on low value data entry type tasks, can put more time and energy towards high value (and more interesting) work.
In short, AI helps eliminate tasks that most employees don’t want to do anyway.
But there is a catch, and that is that AI solutions you integrate into your business need to be user friendly. All the best tech functionality in the world is useless if it’s too difficult or unintuitive to use. So when assessing different options, make sure that user experience is high on your list.
Trovata AI: Built for Finance Teams
At Trovata, we’re in tune with the needs of enterprise level treasurers, working closely with companies like Krispy Kreme, Eventbrite, Etsy, Bosch and more. In doing so, we’ve been able to create Trovata AI, a generative AI assistant designed from the ground up with finance teams in mind.
It allows teams to interrogate and interact with their financial data like never before, with no need to learn code or, in fact, any technical expertise at all.
Teams can run sophisticated scenario plans, build custom models and reports, identify and analyze trends and get instant answers on anything from cash burn to currency holdings to details on recurring transactions.
To find out more about how Trovata AI can help your finance team work smarter, book a demo today.