(Case study) · Financial services · 2025
Rebuilding reconciliation for a forty-year-old lender.
- Client
- NorthLedger
- Services
- Transformation StrategyCustom EngineeringData & AI
- Year
- 2025
(01)
The situation
Month-end close took eleven days and depended on institutional memory rather than systems. Every acquisition added another ledger format, and the audit trail lived in email threads. The bank had been quoted a four-year core replacement. They needed relief in quarters, not presidential terms.
(02)
What we did
Instead of replacing the core, we wrapped it. A matching engine ingests every ledger format the bank has ever acquired, normalises it into a single event stream, and auto-matches with explainable rules; every exception carries its reasoning with it. Deb now trains the rules instead of being the rules.
(03)Measured outcomes
14×
faster month-end close
99.98%
automated match rate
3
legacy systems retired
“They were the first vendor who asked to see our worst spreadsheet before they quoted us.”