Differential privacy in UX data

Differential privacy adds calibrated noise to outputs so individuals cannot be re identified while aggregates remain useful.