The bones of your books, examined and scored.
Before a single figure is interpreted, we assess whether the underlying records can be trusted to carry it. This review measures the structural health of 1 source table — their completeness, their uniqueness, and the small inconsistencies that quietly distort everything downstream.
Completeness & Uniqueness
The share of required cells that carry a value across every table. What is missing is listed below and backfilled where a reliable match exists.
The share of rows that are distinct. Exact duplicates are removed in the cleaned file, with the originals preserved for your record.
Per-Table Integrity
| Source file | Status | Score |
|---|---|---|
ap_export_dirty.csv |
Clean | 99.67/100 |
Structural Findings
ap_export_dirty.csv: 100 missing values detected (amount: 59, invoice_date: 41)
ap_export_dirty.csv: 12 duplicate row(s) found
ap_export_dirty.csv: alias scan skipped for column 'invoice_number' (3,242 distinct values -- over the scan limit for this check to run economically)