Passive Voice Biometrics

for agent calls, IVR, chatbots

The only Voice Biometrics engine to date powered by Machine Learning

Image link
92%
92% - 95%

Authentications completed with agents

20%
10% - 30%

Decrease in AHT

85%
85% - 90%

Authentications completed in IVR

15%
15% - 25%

Increase in Customer Satisfaction (CSAT)

authedication-grey
Authentication with 3 seconds of speech
without a passphrase (entirely “Free Speech”)
Highest-accuracy-grey
Highest accuracy
>92% completion with 3 seconds of speech; >96% with 1-2 additional seconds. Only 4%-8% of all users (inconclusive results) will require additional authentication.
Speech-Recognition-grey
Speech Recognition
for IVR authentications & user enrolment dialogues
quick-deployment-grey
Simple, quick deployment
Background Model generation and calibration is not required.
Self-learning-grey
Self-learning with Machine Learning
Accuracy & performance improves with the amount of processed audios. The more transactions in production, the better the accuracy becomes over time. Human-assisted learning: supervisors can manually add audio to existing voiceprints
Fraud-Detection-grey
Fraud Detection / Prevention and Fraud Analysis
during calls in progress with real-time alerts, and offline investigation on the recordings.
Omni-channel-grey
Omni-channel
the voiceprint of each user can be used in all channels. Only 1 enrollment will be required for each user.
Security-strength-grey
Security strength
results with 99.9% to 99.9999% confidence result. Each Enterprise selects the Acceptance Threshold suitable for its operations.
Passive-ivr-grey
Passive (“Text Independent”)
in IVR, in calls with Agents and also in Mobile Apps, Website and Chatbots. Users can say fixed passphrase (at ABCD my voice is my password ), or they can say their phone number or simply speak freely: “good morning, my name is John” .
Biometric-engine-grey
Full-featured Biometric engine
covering Authentication, Enrollment, Identification (1:N and N:M), Speaker Diarization, Age & Gender Estimation, Language Identification, Speech Quality estimation and Denoiser.
open-reporting-grey
Powerful and Open Reporting
with full capability for the Enterprise users to create custom reports (ad-hoc reports or permanent dashboards in Kibana), and hundreds off-the-shelf integrations to import/export data from/to 3rd party applications.