Her team was producing good work, but reviewing it took time. With nearly a dozen staff, even small errors added up - creating a constant need to check and recheck the numbers.

When we looked at the data more closely, clear patterns began to emerge. Most errors weren’t random - they followed predictable trends, such as misallocations or values sitting far outside the expected range.

With this understanding, I built a tool to automate the review process. Now, data can be pasted into a spreadsheet, and any anomalies are flagged instantly.

This has reduced the time spent reviewing each set of data, saving hours of repetitive work every month - while also making the process more consistent and reliable.