FURKAN SAKIZLIAI DATA ENGINEERING

DATA QUALITY · AI DATA ENGINEERING

Data quality is always purpose-dependent

Data is not abstractly good or bad. Quality emerges from purpose, measurement and risk.

Data points pass through three transparent, purpose-specific quality and ordering frames.
01 · PURPOSEWhich task and decision must the data support?
02 · MEASUREMENTWhich dimension, metric and threshold apply?
03 · RESPONSEWhat happens when the requirement is missed?
CORE METHOD

THE QUALITY PROFILE

Population, dimension, metric, threshold, frequency, consequence and accountability are defined together.

THE SUBGROUP CHECK

Compare the aggregate with at least two relevant subgroups. Test whether the average conceals a risk.