A live caller keeps the signal breathing with its own irregular rhythm. Pipe in a recording and the same line goes mechanically regular — looping, not living.
We don't ask the caller to do anything. We just listen for the things only a live person produces.
The small, uneven pauses and breaths of someone actually talking.
The live acoustics around a real caller, versus a clean captured file.
The natural, non-repeating rhythm a recording can't sustain.
The faint doubled signature left when a captured clip is replayed.
The moments where "is this a real person, right now?" is the whole question.
A recorded passphrase or a cloned reply shouldn't pass as a live customer on the line.
Flag recordings replayed into automated voice prompts and self-service menus.
Confirm a real person is authorizing — in the moment, not after the money moves.
The caller simply speaks. Liveness runs continuously, in real time, and pairs with voice biometrics — biometrics confirms who is speaking, liveness confirms that a live person is speaking right now.
Voice Liveness deploys on-premise. Call audio and analysis stay within your own environment and never leave it.