Before data enters any processing pipeline, I like to do a quick sanity check.
This isn’t about validation rules or automated tests — it’s about catching obvious inconsistencies before they turn into “why is this value weird?” questions later.
In this case, the source data mixed units. Some dimensions were listed in millimeters, others assumed centimeters without stating it explicitly.
Rather than build a temporary script just for this, I manually verified a small sample. Converting millimeter values to centimeters was enough to confirm the data was internally consistent.
I used a lightweight mm-to-cm conversion step (https://mmtocm.net
) as part of that check, then moved on.
Once the data passed the sniff test, it went straight into the normal, automated flow.
Top comments (0)