LedgerlineFintech2022
A treasury dashboard finance teams actually trust
- Role
- Lead product designer
- Timeline
- 9 months
- Team
- 2 PMs, 8 engineers, data analyst
- Platform
- Web app
Outcomes
+38%
Weekly active finance teams
-45%
Reconciliation support tickets
3 days
Faster month-end close
The challenge
Ledgerline's customers loved the data but did not trust the numbers. Controllers exported everything to spreadsheets to double-check balances, so the product was a step in their process rather than the place where the work got done.
Finding out why nobody trusted the numbers
I interviewed 16 controllers and treasury analysts and asked each of them to walk me through their last month-end close. Almost all of them kept a private spreadsheet to “check Ledgerline's homework”.
The problem was never accuracy. It was visibility: when a balance changed, the product never said why.
Designing for explainability
We introduced a single cash position with a running audit trail. Every number can be expanded to show the transactions, rules and people behind it, and every automatic match explains which rule made it.
- A reconciliation queue that sorts exceptions by risk, not by date.
- Plain-language explanations for every automated match.
- Saved views, so each controller sees their own entities first.
Building the system behind it
To ship consistently with eight engineers, I built Ledgerline's first design system: tokens, 46 components and data table patterns, documented in Figma and in code.
Outcome
Within two quarters, weekly active teams grew by 38% and reconciliation tickets fell by 45%. Customers now close their books three days faster on average, and two of the largest accounts retired their shadow spreadsheets for good.
What I learned
Trust is a design problem. Showing the working is often more valuable than polishing the answer.
