IngenID > Use Cases

Our Use Cases

The applications of voice biometric solutions, or use cases, are typically described using 3 elements:

How does your application capture speech? Are you prompting users for short phrases or digits (i.e., text dependent), or are users responding to an Intelligent AI/bot or talking with a live agent (i.e., text independent)? We support both.

Where is the application deployed? We support usage in IVR systems, Intelligent AIs/bots, call centers, mobile apps, enterprise apps, video conferencing services, IoT devices and more.

Why voice biometrics? Are you implementing voice biometrics as part of your multi-factor authentication strategy? Need to detect fraud / account take-overs? Suspect deep fake attacks? Other? We help to address many issues.

Prompting / Speech Collection

Different Ways We Gather Speech

RandomPIN™

Our most popular "prompted" speech collection technique for IVR systems, conversational AI / bots, mobile apps, and more. Users of all age groups are comfortable repeating random numbers, so this approach has wide appeal.

Any Language. We're supporting over 40 languages globally.
Simple Enrollment. Repeat 6 digit phrases one time (< 15 secs of speech).
Simple Verification. Repeat one random 4- or 5-digit number (< 2 secs of speech).
Highly Accurate. An excellent component for multi-factor authentication!

Static Passphrase

Another popular "prompted" speech collection technique for IVR systems, conversational AI / bots, mobile apps, and more. We frequently pair this approach with outbound calls or push messages.

Flexible Prompts. We can support almost any text or numeric prompt of 2-4 seconds.
Use Numbers. For example, repeat your cellphone number 3x to enroll, 1x to verify. Only 10-12 seconds of speech needed.
Use Text. Repeat any passphrase 3x to enroll, 1x to verify. For example: "at Company XYZ, my voice is unique". Only a few seconds of speech needed.
Highly Accurate. An excellent component for multi-factor authentication!

Free Speech / Conversational

This friction-less approach allows you to use regular conversational speech to build voiceprints and quickly very users in real-time. Perfect for conversational AI / bots, live agent conversations, etc.

Any Content. This text-independent or unprompted approach allows your users to just speak.
Enrollment. Robust voiceprints can be built with as little as 30 seconds of speech.
Verification. Just 2-3 second of speech allows you to passively and dynamically authenticate your callers.
Use Recordings. Recordings of fraudsters can be used to create databases of fraudster voiceprints.