A robust, multi-pass reconciliation desktop tool built to automate complex ledger and spreadsheet comparisons across disparate tabular formats (.xlsx, .csv).
Designed for financial auditing, accounting reconciliation, and operations reporting where source files rarely share identical headers, ordering, or exact naming conventions.
- Multi-Format & Multi-Sheet Support: Seamlessly load
.xlsxand.csvfiles, dynamically select individual sheets, and inspect raw headers. - Smart Header Detection: Heuristic-based automatic header row detection capable of ignoring report metadata, banners, and empty rows.
- Configurable Multi-Key Mapping: Align records using up to 3 composite keys with custom fuzzy matching thresholds (50%–100%) or date tolerance windows (
± Ndays). - Multi-Stage Reconciliation Engine:
- Stage 1 (Exact Hash & Fuzzy Multi-Key Alignment): High-speed hash-table indexing for exact records followed by token-set ratio scoring for dirty data.
- Stage 2 (Primary Key Fallback): Re-attempts unmatched records using primary key relaxation.
- Stage 3 (Forced Secondary Fallback): Optional secondary fallback rules for complex, legacy records.
- Math & Variance Computation: Automated numeric delta calculations (detects exact match, rounding drift
≤ 1.0, and absolute differences) across value columns. - Thread-Safe & Non-Blocking UI: Background worker threads keep the Tkinter interface responsive during heavy file processing and scoring routines.
- Interactive Data Preview & Export: In-app tabular results preview with statistical breakdown and single-click export to formatted Excel workbooks.