Are artificial intelligence (AI) and machine learning (ML) worth the hype?
We asked the Head of Innovation at Wolters Kluwer for CCH Tagetik, Fabrizio Tocchini, to shed light on how AI and ML can benefit the finance function, and how you can get the most of these leading-edge technologies.
Letizia Bandoni: One out of three attendees to Gartner’s online event, “How to Make AI a Winning Business Strategy” said they didn’t understand AI benefits and uses. In your experience, does the Office of Finance feel the same?
Fabrizio Tocchini: Well, it depends. The benefits AI can give finance leaders will change significantly depending on how they’re using the AI.
Many solutions on the market allow finance — some requiring the help of data scientists and skilled IT professionals — to perform powerful calculations that apply AI or ML logic. These solutions aren’t ideal. They’re actually quite limiting. Finance gets a hyper-automated number from the technology, delivered out of a black box. In other words, the AI produces a predicted result but provides no context to how it came to its prediction. This “take-it or leave-it” approach is one of the root causes for AI-hesitancy in the Office of Finance.
Finance leaders can’t rely on blind automation. In times of accelerated change, like the one we are living in now, understanding “the why” behind the prediction is a bigger priority than getting a prediction.
This is true even if the prediction is calculated in a very sophisticated, accurate way.
Finance must understand that AI functionality should go beyond predicting what is likely to happen. It should tell you why that prediction is likely to happen. With this knowledge, the Office of Finance can conduct an informed what-if analysis to determine how to best influence future business performance.
Letizia Bandoni: How can the Office of Finance use AI and ML technology to improve financial processes?
Fabrizio Tocchini: To provide value to the Office of Finance, predictive analytics technology has to be explainable.
The intelligent machine must support finance in understanding:
- the what — the predicted outcome
- the why — the drivers determining the predicted outcome
- the how — how to change the predicted outcome
Predictive analytics should provide you with the predicted result — but also information about the drivers influencing the prediction, like a certain product that’s sold through a certain channel.
Once equipped with this information, finance can conduct advanced what-if analysis to “manipulate” the predictions and make informed decisions.
This is what allows predictions to become actionable decisions, consciously taken by empowered finance leaders. Without an explanation, the prediction is one-dimensional. It’s like a body without soul.
I believe this software should be made to be managed by finance. It should be intuitive and immediately usable without requiring technical specialists or data scientists.
Letizia Bandoni: Some finance leaders see this technology as disruptive and don’t perceive it as a priority. Is predictive analytics as disruptive as they think?
Fabrizio Tocchini: Walking across the floors of many finance departments, we noticed there was a lot of fear. No one wants to have to redo the planning process. Many think that AI adoption will impose a steep learning curve and essentially redesign current processes.
Consequently, finance teams perceive adoption as onerous. And we all know from experience that we humans avoid things that are onerous, no matter how good they are for us.
The secret to adoption lies in ensuring a smooth transition with an assisted, sustainable journey.
Many finance teams will be surprised on the amount of hidden insights they can find when they let the “machine” to explore the data they have now.
When you start with what you have now, you can incrementally add in new information later, including operational data and external factors, like macro and microeconomics indexes, customer purchasing information and more. You don’t have to disrupt your way of working.
Letizia Bandoni: What do you mean by “incrementally add in new information?”
Fabrizio Tocchini: When AI isn’t a black box, but explainable and transparent, finance leaders can understand more about their business.
This is where providing the machine with different kinds of data supports the future evolution of AI efficacy for finance. Diverse data and large volumes of data improve the AI logic models and give it more consciousness to interconnect correlations.
By giving AI more detailed operational data, external variables, unstructured information and behavioral analytics, the machine can refine its predictions. In turn, finance’s confidence in the machine improves, as well as the reliability and effectiveness of their business decisions.
Finance can achieve this value by adopting AI in planning process right now.
We are convinced this not a dream anymore. Considering the frequency and the increased volume of data variables we have access to, predictive analytics can be more effective than ever.
Letizia Bandoni: What’s the future value of AI and ML in finance?
Fabrizio Tocchini: AI goes beyond planning and into across all processes. AI will accelerate the closing process, improve reconciliation, facilitate disclosure and regulatory reporting. AI will go beyond finance and improve project reporting across the entire business.
CCH Tagetik expert solutions provide true actionable predictive intelligence. Our budgeting and planning solution enables finance to elevate decision making across many different processes by embedding AI and ML into our platform; it’s an out-of-box capability, not an add-on.
Think about a Tesla. You know you have a next generation car engine when you buy a Tesla.
The same goes for predictive analytics. You shouldn’t have to ask for predictive capabilities. It should underpin your solution. Using this approach, the embedded predictive logic silently revolutionizes your entire business.
That’s the kind of predictive intelligence we’re committed to embedding in the CCH Tagetik platform and we’ll continue to evolve our solution based on feedback we receive from our customers and partners.
After all, we make software to empower people. We don’t make software just to use the technology.