RECORDING
Our panellists discussed current Microsoft capabilities (agents for AP, reconciliation and copilot features), the importance of live connectivity between banks and finance systems, and the practical gaps between theoretical AI functionality and real-world adoption. They emphasized that while ERP tools are expanding AI features, a dedicated TMS often remains necessary for complex treasury tasks like multi-bank connectivity, FX hedging, and debt management.
Our speakers highlighted data quality and integration as the primary enablers for successful AI use cases—cash flow forecasting, process automation, and anomaly detection. Treasurers should prioritize cleaning and structuring data, achieving live bank and ERP connectivity, and designing holistic adoption roadmaps rather than toggling on isolated agents. The panel agreed AI can greatly reduce operational workload and elevate the treasurer’s role toward more strategic advisory work.
The session featured insights from the following lineup of speakers:
- 🎙️Udo Rademakers | Senior Treasurer (Interim Treasurer) & treasuryXL Expert
- 🎙️Theo Wasserberg | Head of UK & Ireland at Embat
- 🎙️Jan-Willem Attevelt | Co-founder of Automation Boutique & treasuryXL Moderator
Key Takeaways
Theo Wasselberg
- “AI can function with whatever data you give it. It can function with last year’s data. Right? But the value, I guess, is in moving from… a dynamic view of what’s happening in the business.”
- “Connectivity is everything in order to feed these models and to get the data into the right place. If you haven’t solved for those connectivities, you are always lagging on the data.”
- “You have to think about how am I going to transform Treasury and roll out AI across it, rather than just waiting for individual agents which have some value in them.”
Udo Rademakers
- “An ERP solely will not do the job, so we need, in Treasury, a TMS, depending on the size of the company… for FX, multi-currencies with hedging, with debt management… I think you need a TMS for that.”
- “Cleaning up the data is really the first thing, and then AI can start working. It’s like human beings. If that cleanup is not in place, you can hardly start with AI.“
- “I think AI can do perhaps a better job than many people can do in cash flow forecasting, and treasury management systems can really help with it, and even detect weird changes straight away.“
Conclusion
The session concluded that AI offers clear practical wins for treasury—especially in forecasting, reconciliation and automating repetitive tasks—but success depends on data readiness and connectivity. Treasurers should focus on pragmatic steps: clean data, connect systems live, and adopt packaged end-to-end solutions that allow agents to interoperate. As AI matures, treasury teams can shift from operational handlers to strategic advisors.
Question to ponder: Which single data/connectivity gap in your treasury would unlock the most AI-driven value if resolved today?





