
Handling languages in Ultimate with the Messaging Widget
This article contains two approaches of collecting language context from customers contacting your AI Agent, powered by Ultimate via the Zendesk Messaging Widget.
Zendesk's native bot supports multiple languages and even offers an auto-translation feature for generative replies and custom flows.
The AI Agent powered by Ultimate uses a slightly different approach in which you can define duplicates of your replies for each language. So where generative replies automatically adapt the language of the conversation, you have full control of the steps and translations used in your flows per language.

My thanks to Cloudset for sponsoring Internal Note
However, in order for this to work correctly you need to make sure you can detect your conversation languages correctly. Zendesk's support help center has extensive documentation on this topic, but in this article I'd like to take a more structured and practical approach to this.
We'll focus on one specific use case, one that is – I think – the most common in use. We have a native Zendesk Messaging widget embedded on a website which is linked to the Ultimate Bot.