Build your agent
Listening
Speech-to-text: the providers that turn the caller’s voice into words, and how to choose one.
| Provider | service | Best for |
|---|---|---|
| Deepgram | deepgram-streaming | Fast, accurate English and Indian English |
| Soniox | soniox | Strong on Hindi–English mixing |
| Sarvam | sarvam | Built for Indian languages |
Getting it right
Add product names, people and jargon to Boost words (keyterm); it's the biggest single fix for mishearing. Set the language callers actually speak rather than leaving it general.