Database Numerical Data Validation

I maintain several live databases that generate a steady stream of purely numerical records—transaction totals, sensor readings, and performance metrics. The raw pulls already sit in secure tables; what I need now is a sharp pair of eyes (and ideally a few smart scripts) to make sure every figure is trustworthy before it feeds our reporting pipeline.

Your focus will be end-to-end validation rather than manual typing. You’ll connect to the databases remotely, run consistency checks, flag outliers, trace discrepancies back to source tables, and document any corrections you make. If you prefer working in SQL, Python (pandas), or even a robust Excel environment linked through ODBC, that is perfect; the key is delivering a clean, fully audited dataset that reconciles against the original record counts.

Deliverables
• A validated export of each assigned table, returned in CSV or the native database format
• An audit log that lists every edit, deletion, or flag you apply, with timestamps and brief reasons
• A short summary report highlighting overall error rates and any recurring data-quality issues discovered

I’ll provide read-only credentials, data dictionaries, and a brief onboarding call so you understand our field definitions. Turnaround on the first batch is one week, and I’m happy to continue on a rolling basis if the collaboration runs smoothly. Let’s make sure the numbers add up.