Ledger

The problem

Founders were exporting CSVs into spreadsheets to understand their runway. The data existed but the insight did not.

The solution

A real-time finance workspace that categorizes transactions, forecasts runway, and surfaces anomalies before they become problems.

How it came together

1. 1

   Modeled a normalized schema for multi-account transactions.

2. 2

   Built streaming categorization with a human-in-the-loop review.

3. 3

   Designed dashboards around decisions, not raw numbers.

Outcome

• Monthly close time reduced from days to minutes.
• Anomaly detection caught real billing errors in week one.
• Adopted by every team member, not just finance.