Skip to content

Audio: voice messages

It allows your assistant to listen to customers' voice messages and respond with voice. Use it to make communication more natural on channels like WhatsApp and Telegram.

How it works

The system listens to the audio file and transforms it into text to understand the request. For example, if a customer sends a voice message asking: "Hi, what time do you close tomorrow?", the assistant understands the question and responds with the correct time. If the channel allows it, the assistant responds with a voice message instead of text.

You can manage how the assistant responds by checking the personalization section.

What you need to do

Nothing: the service is automatically active on channels that support audio.

What glacom® does

glacom.AI receives the audio file, transcribes it in real time and analyzes the content using the information uploaded in knowledge. If planned, it generates a natural voice response for the user.

Limits

The system may have difficulties in these cases: - Very loud background noise that covers the voice. - Very strong dialects or extremely marked accents. - Excessively long voice messages.

Frequent problems

  • Did the assistant misunderstand a word? → Check the transcription in conversations and use errors and corrections to refine the response.
  • The customer speaks a dialect and the system doesn't understand? → In this case the assistant might ask for clarification or you can intervene via operator.
  • Can it be used on every social network? → Only on channels that provide an API compatible with sending and receiving audio, as explained in integrations.

From the app

You can listen to the voice messages received from customers and read the generated text transcription.