Reaching the right party

Why we built it

Verification decides whether a call can go anywhere. Before an agent can discuss a balance, it has to confirm it is speaking to the right person, and on a phone line that depends on hearing a name and an address correctly.

Telephone audio makes that hard. Speech recognition regularly mishears proper nouns, and Spanish, and other non-English names suffer most. Matching names as text made it worse. A strict check rejected Erica because the record said Erika. A loose one let a partial match through. Either way the call went wrong: the right person was turned away, or someone else got close to hearing details that weren't theirs.

How it works

proAgent's AI now verifies names by how they sound, the way a person on the phone would.

• Before the call starts, the consumer's name, address and email go to speech recognition as hints, so those words are heard correctly.
• When details arrive later in the call, from an account or phone lookup, the hints update without interrupting the conversation.
• Names are compared by sound and phonetic spelling, so John and Jon match while a different name with a similar shape does not.
• When the first name matches and the last name is unclear, the agent asks for the last name again instead of guessing.
• Hints can also be set for each stage of the call, so the words that matter at that point are the ones recognition listens for.

What you get

More of the right consumers clear verification and reach the conversation about their account, instead of being turned away over a spelling or an accent. The matching stays strict enough that the wrong John Smith doesn't hear someone else's balance, which matters as much for compliance as it does for conversion.

Sound-based name verification is available in proAgent's built-in verification flow. Talk to your Prodigal team member about enabling it on your agent.