The Treasury Time Traveller
The Situation
When a fast-growing B2B SaaS company expanded across Europe and the US, its treasury needs grew with it. What had once been handled alongside accounting was becoming too complex to manage without a dedicated treasury function. Different entities had their own banks, payment service providers, systems, and processes – leaving plenty of room for consolidation and improvement.
The company’s Treasury Manager had already built the treasury function from the ground up and introduced a central platform for cash visibility and payment execution.
But one major challenge remained: cash flow forecasting.
The existing process was largely spreadsheet-based, requiring significant manual preparation before treasury could even begin analysing the numbers. For a small team supporting a rapidly scaling organisation, that simply wasn’t sustainable.
The Challenge
Essentially, the problem was turning all of the useful, existing data into usable. Information sat across different systems, structures and spreadsheets. Preparing actuals, categorising cash flows and comparing forecasts against reality required exports, corrections and reformatting. By the time treasury reached the interesting part, that is, understanding what the numbers actually meant, a significant amount of time had already been spent preparing them.
And as the business expanded into new markets, improved its banking and PSP infrastructure and began M&A activity, the need for flexible and timely forecasting only increased.
A traditional TMS seemed like the obvious next step, yet in this case, was it necessary?
The company already had cash visibility and payment execution in place. Starting another large connectivity and implementation project would have added cost, complexity, and months of work.
What treasury needed was much more focused than a traditional TMS :
A flexible forecasting engine that could work with the existing setup, automate the preparation layer and help a small team move from data to decision much faster.
The Approach
Rather than following the traditional route of implementing a full TMS, the Treasury Time traveler focused on the specific gap that needed solving.
With support from Pecunia and discussions around the available treasury technology landscape, the search narrowed towards a specialised AI-powered cash flow forecasting solution.
AI in this context was the opportunity to work smarter while keeping treasury lean.
The selected solution could automate data inflows, structure and categorise actuals, support forecast creation and make actual-versus-forecast analysis much easier to access. Implementation took a couple of months and was deliberately kept focused.
One challenge was bringing actual cash-flow data into the tool without rebuilding connectivity with every bank and PSP. Another was historical data: because treasury itself was relatively new, there was not yet years of perfectly consolidated information available for the AI model to learn from.
Rather than waiting for perfect data, the team took a pragmatic approach.
Start with what is available -> Build the right structure->Refine as more history develops.
AI became the support layer, while treasury remained firmly responsible for assumptions, review, interpretation and final decision-making.
The Impact
The biggest change is simple:
Treasury gets to the meaningful work faster.
Instead of spending large amounts of time preparing and cleaning data, the team can increasingly focus on analysing results, challenging assumptions and understanding what action needs to follow.
Forecasting has become more dynamic, shortening the distance between data and decision. Manual work has also been reduced significantly (an especially important benefit for a small treasury team).
But the value goes beyond efficiency…
By removing much of the preparation layer, treasury gains more capacity for the work that previously had to wait: investigating trends, exploring opportunities, refining processes, and contributing more strategically to the business.
The solution is still relatively new, so its full impact will develop over time as more actuals and forecast history become available.
But the direction is already clear:
AI isn’t replacing treasury judgement. It is creating more room for it.
And that is a use case example of a very important message we emphasize at Pecunia; check out one of our latest blogs on the topic HERE
From Forecasting the Future to Shaping It
For growing treasury teams, the answer isn’t always a bigger system or a bigger team. Sometimes it is finding the right technology for the exact problem in front of you.
Pecunia helps treasury teams identify the right solutions, challenge the obvious next step, and build a treasury setup that can grow with the business.
Ready to spend less time preparing the forecast and more time acting on it? Let’s talk.
