Despite its transformative potential, some huge misconceptions exist around AI. It’s often seen as a cure-all - a technology that can be deployed in a blanket manner across the finance function to instantly add a layer of sophistication like an all encompassing umbrella.

But, in reality, AI requires a much more focused and nuanced approach to deliver meaningful improvements and ROI.

Specialized digital agents

Rather than relying on a single AI system to do everything, successful finance teams deploy AI in selected processes with a laser focus on particular goals. And cutting-edge finance functions are starting to use specialised AI agents (in some ways similar to digital assistants like Siri or Alexa) to achieve targeted results.

For example, one AI agent might focus on hedging and FX, while another tackles payments, and a third works on investment optimization. Cash flow forecasting is, of course, also ripe to have an agent of its own.

These agents process data from a variety of internal and external sources, including ERPs and banks. But they also actively monitor for risks and opportunities. And, over time, each agent will learn your company’s specific patterns and preferences, providing increasingly sophisticated support.

When reviewing thousands of payment transactions, for instance, an AI agent can flag anomalies that might indicate fraud or unusual activity. And for bank fee analysis it can automatically identify overcharges or opportunities for fee reduction. 

In investment management, the agent can spot yield optimization opportunities across different currencies and instruments. For cash forecasting, the agent can predict cash needs based on historical patterns, recommend adjustments to payment terms, and even identify potential shortfalls before they occur. And the list goes on.

The human element

An AI agent doesn’t replace the need for human expertise. Instead, it acts as a decision support system, enhancing the existing abilities of finance professionals. 

So, for example, while AI can flag a potentially suspicious transaction, it should still be up to the finance team to assess the context and determine the appropriate action. The new mantra is ‘human + AI’ or ‘human in-the-loop’. And this collaboration works particularly well in areas like:

• Working capital optimization: AI identifies opportunities to improve liquidity management which the team can then assess and, where appropriate, act on.
• ‍Bank fee analysis: AI flags discrepancies in fee structures across accounts. But the team will then need to leverage their negotiation and communication skills to work out a better deal from their banking partner, while keeping the relationship intact (and beneficial for both parties).
• Yield management: The AI agent might suggest better ways to allocate surplus cash, within pre-defined treasury policy limits. But AI should not be acting on these suggestions alone. The team should still be making the decisions based on their own input, experience, and expertise and the broader business objectives.
• ‍FX hedging: Market conditions can easily be evaluated by the AI agent to help inform hedging decisions. And while tech can even make suggestions as to strategies to deploy, it is ultimately down to human team members to make the final call.

A new way of working

The beauty of AI-assisted finance functions is that small finance teams can now operate with the efficiency of functions twice their size. That might sound like a bold idea or grand ambition, but transitioning to AI-assisted operations doesn’t have to be overwhelming. 

Successful implementations often start small, focusing on establishing connectivity and visibility. Once teams have a clear view of their cash positions and transactions, they can begin leveraging AI for more specific use cases. This kind of phased approach builds confidence in the technology while ensuring that critical processes remain under control. 

Think of it like onboarding a new team member on  probation.. Initially you may only rely on them for an hour or so a day.. T But as trust grows, and they better understand you and your business needs, their involvement can scale.

Adapting to this new way of working will require some time, training, and trust-building. But the payoff is clear: smaller teams achieving greater impact, professionals freed to focus on strategic decisions, and finance functions that are more responsive and resilient than ever before. So, perhaps the question is not whether you will recruit an AI agent to your finance team in 2025, but when. 

Curious about how others are putting AI to use? Let’s chat  – get in touch with our team to explore the possibilities.