Expense management automation that gave finance a working day back
A finance team got back a working day every month, while expense data that had been manually gathered from multiple sources became continuously available and automatically updated.
The situation
For a company with multiple remote teams, employee expenses and their supporting data lived across several platforms. Once a month, a member of the finance team exported data from each source, cleaned and reformatted it, reconciled the differences between platforms, standardised categories, and combined it into a single reporting structure.
It took roughly a working day. Until the reconciliation was complete, nobody had a current picture. Gaps and inconsistencies were hard to catch because they surfaced only in the process intended to resolve them.
The constraint
Moving records between systems was the easy part. The harder problem was building one unified data model from sources that represented expenses — and sometimes employee names — differently, while keeping it resilient as the company changed.
Some differences appeared only once real data began flowing. What the platforms exposed programmatically did not always match the format finance had received through manual exports, so those differences had to be found and accounted for during the build.
The rest was structural. Categories, employee records, departments, and the mappings between them could not be hard-coded and forgotten. New people join, departments change, and finance introduces new cost categories as the business evolves.
The system therefore had to automate the current process without creating a new dependency every time the underlying structure moved. Finance can maintain category and department mappings directly in the system rather than requesting an engineering change for every update.
What we built
An expense-data pipeline in n8n. Records are collected as they appear where a platform provides webhook events, and on a schedule where it does not. Original records are retained as received, while a separate unified layer cleans, maps, and enriches them into the structure finance reports on.
When an expense is shared between multiple employees, it is attributed across everyone involved rather than landing entirely on the person who made the booking. Employee data also stays in sync automatically.
Process diagram
Where it stands
The system is live and stable, with tens of thousands of records processed. The main build is complete and the pipeline runs without regular intervention. Swiss Product Studio remains involved when the team wants to extend the automation.
Next
AI quality control: what automated visual inspection would really take
Ready to boost business results with AI automation?
Book a 30-minute operations review. We will look at where manual work is limiting your operation and whether there is a practical opportunity to create capacity.
Book a 30-minute operations reviewSwiss precision. Valley speed.