Set up multiple versions of your agent so that it is ready to speak more than one language.
Currently, Agent Studio supports only one language per project. If you want to support multiple languages, we recommend creating a separate project for each language and localising the relevant content accordingly.
When configuring a project for a language other than English:
multilingual_v2
model.
Ask the user for their number by saying “¿Me puedes dar tu número de teléfono?”
Some project content needs to be translated, and some does not. Here’s a breakdown:
Area | Element | Yes | No | Notes | ||
---|---|---|---|---|---|---|
Knowledge Base | Sample questions | Must match user input language for retrieval | ||||
Content | Translate for brand accuracy and better output | |||||
Topic names and actions | Keep in English (used internally, not user-facing) | |||||
SMS | SMS content | Translate anything user-facing | ||||
ASR & Voice | ASR keywords and corrections | Leave in native language. These might be very different to the English configuration | ||||
Response control and pronunciations | Leave in native language. These might be very different to the English configuration | |||||
Functions | Python code | Leave in English | ||||
Function names and descriptions | Leave in English | |||||
Hard-coded responses and LLM prompts | Translate only user-facing content (e.g., utterances) | Translate only user-facing content (e.g., utterances) |
If you’re using a function like:
That string should be translated.
If you’re re-prompting the LLM:
You only need to translate if you’re giving an example response.
Prompts are found in:
If you have questions about setting up multi-lingual support or want to share your team’s approach, please reach out to platform-support@poly-ai.com.
Set up multiple versions of your agent so that it is ready to speak more than one language.
Currently, Agent Studio supports only one language per project. If you want to support multiple languages, we recommend creating a separate project for each language and localising the relevant content accordingly.
When configuring a project for a language other than English:
multilingual_v2
model.
Ask the user for their number by saying “¿Me puedes dar tu número de teléfono?”
Some project content needs to be translated, and some does not. Here’s a breakdown:
Area | Element | Yes | No | Notes | ||
---|---|---|---|---|---|---|
Knowledge Base | Sample questions | Must match user input language for retrieval | ||||
Content | Translate for brand accuracy and better output | |||||
Topic names and actions | Keep in English (used internally, not user-facing) | |||||
SMS | SMS content | Translate anything user-facing | ||||
ASR & Voice | ASR keywords and corrections | Leave in native language. These might be very different to the English configuration | ||||
Response control and pronunciations | Leave in native language. These might be very different to the English configuration | |||||
Functions | Python code | Leave in English | ||||
Function names and descriptions | Leave in English | |||||
Hard-coded responses and LLM prompts | Translate only user-facing content (e.g., utterances) | Translate only user-facing content (e.g., utterances) |
If you’re using a function like:
That string should be translated.
If you’re re-prompting the LLM:
You only need to translate if you’re giving an example response.
Prompts are found in:
If you have questions about setting up multi-lingual support or want to share your team’s approach, please reach out to platform-support@poly-ai.com.