Language Transparency for Agent Handover

During internal discussions, we explored the possibility of enhancing agent awareness in multilingual support scenarios. The idea is to implement a mechanism that notifies the agent when a user's message has been translated by a Large Language Model (LLM). This could be achieved by:

• Displaying a subtle note in the chat history, such as:
  “This message was translated from [Original Language] to English via LLM.”
• Send the original message in the chat history.

This feature would help agents better understand the context and potential nuances of the conversation, improving empathy and response accuracy.

For eg: In certain language translation the keyword “outlook” gets translated to “view”. This note enhances transparency for agents by indicating when a message comes from a non-native English speaker.