Foundations First: Treasury Transformation Starts With Data, Not AI
Sanae Maamouri, liquidity and investment operations director at Invesco and board member of ATEL, explains why treasury transformation begins with data, not AI. Centralise your bank and account data, understand your real needs, then automate, because AI amplifies whatever you feed it. Her message to treasurers: stop entering figures, let the system generate them, and move from executing to analysing, recommending and deciding. As she puts it, “we are not replacing the human, we are just removing those non-value added tasks.
An interview with Sanae Maamouri, liquidity and investment operations director at Invesco and board member of ATEL.
“We are not replacing the human, we are just removing those non-value added tasks.”
Sanae Maamouri, Invesco
What key items should treasurers think about for 2027 and 2028?
Everyone is talking about AI and everyone wants AI. Everyone wants to automate as much as possible of treasury tasks. But before all that, we should think about the foundation. When I talk with peers, I see companies without even centralised data on their banks and accounts, so they cannot go forward. What do you want to transform? You should first build a strong foundation and all the necessary databases, structure your work and understand your needs, because it depends on the structure. For some, the priority lies in a precise, reliable forecast. For others, visibility remains the most urgent, because they do not even know where the cash sits at a central level. Each company should understand what happens inside and start from there. The database forms the first foundation before all these trendy topics.
How is your company replying to these requirements?
The feeling is positive, because I work with people who welcome change. Building the data foundation proved easy, since everyone wanted to participate and reach the final outcome: easy access to cash balances, tracking of all kinds of transactions, forecasts generated at any moment, and a move from only analysing to getting profit from treasury by investing and optimising cash. That mindset made the transformation a successful journey. Five years ago, people feared change more, with the idea that automation and new tools would replace the human, which is not true. We are not replacing the job, we are removing non-value added tasks and bringing a more analytical mindset. We tell treasurers: stop entering figures, the system generates them, you analyse and recommend. You no longer execute, you lead and take the decisions.
What key challenge should people keep an eye on going forward?
Before moving to AI, be sure about what you feed your AI, because whatever you feed, it amplifies. Feed bad data and it multiplies the errors. Feed quality data from a strong source foundation and you get very good results. Take a basic treasury example. If the AI consolidates my positions and generates my short-term forecast based on a bank account that still shows a balance but is in reality closed, my forecast relies on cash that no longer exists. If I expect to pay a maturity that day from that account, I will face a delay and a penalty. AI remains a nice addition to structured work and a strong foundation, never a replacement. Used wisely, it can only benefit you, your customers and all your stakeholders.