Talk to your AI Agent via voice or phone

Speaking is the most natural forms of human interaction. With Cognigy.AI, you can enable your AI Agent to engage through voice communication. By leveraging endpoints like the Cognigy Voice Gateway, you can integrate voice features or connect to phone systems seamlessly. For the best outcomes, you can also fine-tune the AI Agent's responses to align with your needs.

We recommend to activate Streaming for lowest latency responses. AI Agents support configurable, optimized phrase-by-phrase streaming of responses to reduce latency. To enable this feature, navigate to the Storage & Streaming Options section in the AI Agent Node settings and set your AI Agent to "Stream to Output".

To test the voice experience in Cognigy.AI and if you don’t have Cognigy Voice Gateway configured, use the Voice Call tab at the bottom of the Interaction Panel by following these steps:

1. Set up a Speech Provider.
   To enable the voice preview, configure a Speech Provider such as Amazon Polly, Google Speech Services, or Microsoft Azure Speech Services. You can do this via the sidebar under Settings > Voice Preview Settings. Once configured, the phone icon in the Voice Call tab of the Interaction Panel will be activated.
   
   
2. Configure Language and Voice.
   Add a Set Session Config node at the beginning of your Flow to select the Text-to-Speech (TTS) and Speech-to-Text (STT) vendor based on your chosen Speech Provider. Use this node to specify the TTS and STT language and select the STT voice for speech synthesis.

Finally, click the phone button in the Voice Call tab of the Interaction Panel to interact with your AI Agent using voice.

➡️ To get your AI Agent ready for prime time, let's wrap it up with a look at debugging.