We spent a year interviewing more than 400 CFOs, controllers, and finance leads across Europe before we wrote a single line of code. The pattern was impossible to miss.
What 400 conversations kept surfacing
Finance teams told us the same thing in different words. Pre-accounting, the work that happens between an invoice landing in an inbox and a clean entry reaching the ledger, was eating their time. Roughly 20 minutes per invoice. Across 500 invoices a month. That works out to more than €40,000 a year in team hours on work that is mostly repetitive logic.
The striking part was not the number. It was how normal everyone treated it. Nobody framed it as a problem to solve. They framed it as the cost of doing business.
What we were going to build, and why we did not
Our original thesis was an AI-native FP&A platform. Strategic forecasting, variance analysis, scenario modelling. We were convinced that is where finance teams needed help.
The interviews changed our minds. Strategic tools are only useful when the operational layer is stable. Every CFO we spoke with who was drowning in dashboards was also drowning in unprocessed invoices. You cannot build a clear view of the future on a messy present.
"There is a lot of information in finance that you simply cannot derive automatically from the system pathway today."
Why agents, not dashboards
Pre-accounting does not need more visualization. It needs execution. The work is high-volume, rule-based, and full of exceptions that require traceable decisions. That is exactly what agent-based reasoning handles well.
Our agents reach around 90% automation within seven days on real company data. The remaining exceptions are surfaced with the reasoning attached, so a human can review and close them in seconds rather than minutes. Nobody does the work in between for free. We wanted to change that.
How we could afford a year of listening
Operating within Builders Studio gave us the runway to research before building. The co-founder personally ran 400 calls with finance leaders. We were not pitching. We were listening. The time investment paid off because the product we shipped is the product people described needing, not the one we assumed they wanted.
The signal that told us we were right
A Rabobank Innovation Loan came in before we had shipped a complete product. External validation of the research and architecture approach. It told us the pattern we found was real, not just an artefact of who we happened to call.
What we are committing to
Cortena is a multi-year initiative, not a quick exit. We start with payables because that is where the volume and the pain are highest. From there we move into reconciliation, spend management, receivables, and eventually the strategic layer we always wanted to build. The difference is that now we know the order matters.