U.S. Bank is developing AI-driven solutions for its business clients, with its latest efforts targeting the launch of an AI-driven cash-forecasting tool. 

Major financial institutions like Bank of America, Citizens Bank and PNC Bank have already deployed cash-forecasting solutions.

The $669 billion bank is working with cloud-based fintech Kyriba on the tool, which will help businesses better manage finances, Alberto Casas, head of product for global treasury management at U.S. Bank, told Bank Automation News. 

“We look forward to introducing a highly automated cash positioning and forecasting solution that provides clients with intelligent and comprehensive views of their cash and liquidity positions,” Casas said. “Our key priorities are to provide better visibility of their future cash flows, reduce manual work processes related to gathering and generating financial forecasts, and offer easy access through our banking portal that they already use every day.” 

Kyriba’s platform assesses a business’s historical payment trends, cash balances, predicted inflows and confirmed outflows of cash. External data factors like fluctuations in currencies are also assessed to provide users with a tentative image of how their cash balances might look in the future, according to Kyriba. 

The tool is expected to launch by the end of the year or early 2025 as a larger effort of the bank to develop and deploy technology for its customers, Casas told BAN. 

In the second quarter, the Minneapolis-based bank spent $509 million on technology and communications, up 2% year over year, according to the bank’s July 17 earnings report. 

The macro 

As interest rates remain elevated, many businesses are looking within to manage their working capital and search for short-term loans, Shruti Patel, chief product officer for business banking at U.S. Bank, previously told BAN. The change in the market cycle and new technology like AI has prompted the bank to develop tools like cash flow forecasting, Patel added. 

However, developing a cash forecasting tool is difficult, Liran Zelkha, co-founder of SMB business banking fintech Lili, told BAN. 

“Everybody wants to do it, we want to do it, but the truth of the matter is that [creating good cash forecasting] has been tried for years,” he said, adding that accurate forecasting is “still incredibly hard” to do. 

The problem is not that AI tools aren’t good enough but that “bringing in all the data in a good enough way,” is difficult, Zelkha said. 

For cash forecasting to work accurately, external data is as important as a company’s internal data, which can be challenging to add to an AI model, he said. 

Early-bird registration is now available for the inaugural Bank Automation Summit Europe in Frankfurt, Germany, on Oct. 7-8! Discover the latest advancements in AI and automation in banking. Register here and apply to speak here.