Call audio capture
Capture call audio through SIP and an external capture module for modeling and comparison.
Solutions /Telephone Voice Verification
Add an identity check during the call.
Use existing calls for identity verification. Compare captured speech with registered voiceprints to assist call center identity checks, while managing speech, models, and call records centrally.
Product features
From audio capture
to in-call identity comparison.
Customer voiceprint modeling
Build voiceprint models from captured speech and store them for verification in subsequent calls.
In-call voiceprint comparison
Compare current call audio with registered models and provide recognition results during the conversation.
Call and model management
Manage captured speech, voiceprint models, verification information, and call records in one place.
System architecture
Deployment
that fits your business.
Review SIP integration, audio capture, and customer systems, then choose an add-on or standalone deployment according to network conditions and call volume.
Existing systems
Review business systems and integration needs
Network isolation
Define network conditions and isolation requirements
Business scale
Evaluate capacity and expansion requirements
Advantages
Keep your phone services.
Add voiceprint capabilities.
Two deployment options
Add a voiceprint module to an existing call center or deploy a standalone telephone voiceprint system.
Two verification approaches
Use text-independent or dynamic-passphrase verification according to the scenario.
Use existing recordings
Create models from call recordings or use historical call audio for offline modeling.
System integration support
Custom development connects to other business or biometric verification systems.
Workflow
Build models first.
Compare voices in later calls.
- 1
Prepare speech and create models
An unregistered person's first call produces no recognition result. Use that recording to register a voiceprint afterward, or model from historical audio offline.
- 2
Provide results during calls
During subsequent calls, capture and analyze speech against registered models to provide identity recognition results.
