FAQ
Common Reconify setup and reconciliation questions.
Which file formats are supported?
CSV, JSON, NDJSON, XLSX, and XLSM are supported. Legacy .xls files are not supported.
Why are amounts large integers?
Reconify stores amounts in minor units. For a two-decimal currency, 12345 means 123.45.
Which side should be left or right?
The right side is indexed first. For large runs, choose a right file that fits the desired index backend. Behaviorally, left and right also determine which unmatched section a row appears in.
Why did a row become amount_diff?
The reference matched and the date was within the window, but the amount exceeded amount_tolerance_minor.
Why did a row become timing_diff?
The reference and amount matched, but the date difference exceeded date_window.
Why do I have many unmatched rows?
Check parser fields first. Wrong date, amount, reference, or name columns can cause expected matches to miss. Run config check-source and parse before changing matching rules.
Does Reconify support fuzzy matching?
The implemented fallback is token-based name matching with name_mode: "tokens". Reference matching always runs first.
Can I run very large files?
Yes. Use streaming output (ndjson, csv, or json-stream), configure index.backend: auto or disk, and enable --progress.
Can I get audit metadata?
Yes. Use --audit with json, json-stream, or ndjson.