Today's definition of a customer-centric business model is incomplete without automatic speech processing components which aim to serve the clients’ needs efficiently. However, deploying these components results in large-scale voice data collection containing private, sensitive, and personally identifiable information about the speakers. European privacy laws such as the GDPR require data anonymization and removal of the biometric information from the collected speech data. This talk explains the voice anonymization platform Nijta, which quickly removes speaker’s personally identifiable information from their voice data, and says that the process of anonymization grants wide access of this data to the organisations while allowing them to have trustful GDPR-compliant user interaction with their customers.
