Treasury Services Is Entering Its AI and Automation Era
More money brings more problems, as the saying goes. It is something that the treasury and finance functions of mid-size and large enterprises know well.
After all, the larger the scale and greater the diversity of a firm’s operations, the larger the scale and the greater diversity of their cash management workflows — and this complexity can hinder efficiency, increase risks and create challenges in maintaining optimal liquidity.
“Cash flow can be a blind spot for the finance team,” Noam Mills, CEO at Panax, told PYMNTS, explaining that traditional cash flow management can often be reliant on manual processes and reactive measures.
She noted, drawing on her own experience within the finance function at a global eCommerce business, that treasury teams are often “chasing their own tails just to understand where they are and to make decisions.”
But with the rise of artificial intelligence (AI), modern solutions leveraging the innovation are increasingly playing a pivotal role in automating and streamlining financial processes for complex treasury organizations.
“The key word here is complexity. And complexity can arise from many different sources, whether it’s the holding structure or the nature of the business,” said Mills.
She noted that pain points can span from the realities of businesses working across multiple geographies and dealing with different currencies; to others whose pain stems from a multitude of accounts, even in one geography, that can make managing working capital a challenge. Operational and treasury-centric complexities frequently also arise from the usage of credit lines on one hand or interest-bearing account on the other hand.
“All these things create the need for more proactive management,” said Mills.
And mitigating these traditional headaches with AI can allow for more proactive cash management, significantly enhancing the ability of finance teams to make informed decisions swiftly.
The Role of AI in Transforming Cash Management
The benefits of artificial intelligence in cash management are manifold. By automating low-level tasks, AI frees finance professionals to focus on strategic decision-making. For instance, Panax uses AI to categorize bank transactions more effectively, ensuring that finance teams have accurate data to base their decisions on. This not only improves efficiency but also enhances the accuracy of financial forecasting and liquidity management.
This shift, Mills said, is akin to moving from using printed maps to leveraging dynamic GPS systems like Google Maps or Waze.
“Proactive AI solutions are helping even lean finance teams that don’t necessarily have the manpower of large treasury teams to be in control and make more optimal decisions,” she said. “It’s a tectonic shift that is happening rapidly.”
One key enabler of these advancements in automating routine tasks and providing real-time insights is open banking, which facilitates secure and real-time access to financial data.
Without such infrastructure, the development of sophisticated AI-driven solutions would be much more challenging, and Mills emphasized that open banking, despite its slow pace of advancement, has been playing a crucial role in providing the necessary data connectivity that AI algorithms depend on.
Enabling the Future of the Finance Function
Looking ahead, Mills envisions a future where AI-driven cash management platforms operate almost autonomously. Finance teams would set policies and guardrails, while AI handled execution, akin to a co-pilot managing the technical aspects of flight. This would allow companies to optimize liquidity, minimize risks and maintain full control over their financial operations.
“There is a real paradigm shift from like older solutions that focus more on consolidating the data and function as a reactive platform that still requires the user to monitor, analyze, and make decisions at all levels to a proactive product that serves as a co-pilot,” Mills said.
Despite the enthusiasm surrounding AI, Mills acknowledged its limitations and challenges. Data quality remains a significant hurdle, with AI models only as good as the data they are trained on and poor data quality frequently leading to ineffective solutions.
“We are dealing with complex decision making in companies that have a lot of complexity,” Mills said, stressing the importance of embracing high-value solutions that have both trust and credibility.
