Skip to content
Start a projecttafseel@walqalum.com

Work · Case study

Reconciliation thatcloses the month on time.

Client
Finance team, Pakistan
Year
2025
Scope
Matching engine, exception queue
Stack
ML · Pipelines · Dashboards

01 / Challenge

Month-end reconciliation was a spreadsheet exercise. Thousands of lines were matched by eye, and the exceptions that actually needed judgement were buried under the ones that did not.

02 / Approach

We built a matching engine that clears the obvious pairs automatically and scores the rest by how uncertain they are. Analysts see a queue ordered by what needs a decision, not by what happened to be next in the file.

Every automatic match is reversible and logged with the rule that made it, so an auditor can follow any line back to its evidence.

03 / Outcome

92%

Matched automatically

4 days

Off the close

The close runs on the exceptions rather than on the whole ledger, and the analysts spend their month-end on the entries that need them.

Next project

Portfolio Valuation Support
Start a project