Building a data-quality gate for 5,000 records.
Of 5,000 records, 4,090 passed validation: an 81.8 percent data quality rate.
What does the work involve?
Portfolio case study. All data is fictional and simulated.
Problem. A customer transaction dataset arrives with no quality control: blank required fields, duplicate IDs and transactions, malformed emails and phone numbers, and values outside the reference lists.
Approach. I built a validation system in Microsoft Excel, native features only: 13 checks per record, all driven by formulas, dropdown Data Validation at the entry point, status flags shown in colour, and a pipeline of 8 sheets from raw input to a clean, standardized dataset. A QA_Check sheet runs 11 automated checks and every number on the Summary dashboard recalculates live.
Result. Of 5,000 records, 4,090 passed validation for an 81.8 percent data quality rate, with 11 out of 11 QA checks passing. The workbook is fully reproducible from a seeded generator script, so every figure can be rebuilt and re-verified.