01
Problem
Holding-wide reporting and loan schedules were assembled manually, while early AI expense-classification experiments were not reliable enough for finance.
A loan calculator, holding dashboard and reporting workflows gave finance a clearer operating system for recurring work.
What was blocked, what changed, and what the client gained.
01
Holding-wide reporting and loan schedules were assembled manually, while early AI expense-classification experiments were not reliable enough for finance.
02
We built a public loan calculator, a portfolio dashboard and reusable reporting workflows, including side-by-side evaluation of AI tools for a real task.
03
The team gained a single financial view and repeatable reporting tools instead of relying on separate files and manual table-building.
Financial reporting across the holding was consolidated manually, so every figure needed to be checked again. Loan calculations and repayment schedules also lived in spreadsheets, and early AI expense-classification experiments produced too many errors to trust in a finance workflow.
Finance has a clearer portfolio-wide summary, a ready calculation tool and a repeatable approach to assembling recurring reports. The solution avoids unverified claims while measurable production outcomes are still being tracked.
Let's scope the system around it.