Completed project02 / 10

Spreadsheet rescue

Data rescue / inconsistent data becomes controlled, reviewable input

An Excel cleanup and formula-repair system that separates uncertain records instead of hiding them.

Live project viewData cleanup / QA console18,420 rows checked
Before / exceptionsINV-008112/7/26Vendor A$1,8O0INV-008107-12-26VENDOR A#VALUE!Missing2026/07/13Vendor A$920
98.7%ready124 exceptions isolated
After / controlled outputINV-00812026-07-12Vendor A$1,800INV-00822026-07-13Vendor A$920INV-00832026-07-14Vendor B$640
DuplicatesFormula traceType repairQA export

The challenge

A business-critical workbook contains inconsistent dates, duplicate records, broken formulas, and values that look right but cannot be trusted.

01 / Approach
  1. Profile the workbook before changing it
  2. Trace formulas and standardize data types
  3. Separate safe corrections from human-review exceptions
  4. Deliver before-and-after QA evidence
02 / Deliverables
  • Cleaned workbook
  • Formula map
  • Duplicate controls
  • Exception queue
  • QA summary
  • Repeatable cleanup routine

What the finished solution gives the team

Visible logicImportant transformations and assumptions can be inspected.

Exception pathsUncertain inputs are surfaced instead of silently forced through.

Operable handoffThe owner receives a system they can understand and run.

Next projectAudio workbench / long recordings become a precise, searchable clip library