Once a speaker (customer) is transferred to a contact center agent and starts to speak, PerSay FreeSpeech listens in to the conversation and retrieves the speaker's audio for processing. The acquired audio is compared to the speaker’s voiceprint stored in the system, and a verification result is generated within seconds. The verification result is then transferred to the CTI, agent workstation or CRM system. As the conversation proceeds and more audio is acquired, further verification results can be generated.

Enrollment

Enrollment is performed while the user is engaged in a normal conversation with a contact center agent. The system records the first one or two conversations with the user, extracts the voice features that are unique to the specific person, automatically creates a voiceprint, and stores it in a secure directory.

Verification

Verification is performed in real time in the background of a natural conversation. Captured samples are compared to the speaker's stored voiceprint; a verification score and decision are available within 5-10 seconds.