Technologies for privately processing voice data include a compute device configured to continually or periodically capture voice data of a user by the compute device. The captured voice data is processed to remove or reduce the user's privacy-sensitive information. For example, the compute device fragments the captured voice data to generate a set of voice data fragments and further scrambles the voice data fragments to generate scrambled voice data fragments having a sequential order different from the plurality of voice data fragments. To scramble the voice data fragments, the compute device may reorder the voice data fragments such that each fragment is repositioned from its corresponding original sequential position in the voice data by a particular number of words, syllables, or phrases.