This feature is in Early Access release.
Introduction
To make sure consumers get the most accurate and helpful answers, intelligent systems often start by carefully understanding the question or query. The initial processing of the user’s query plays a vital role and can involve using an LLM.
How it works
Custom processing of the user’s query refines and prepares the input before it’s used to search the knowledge base. This ensures that the search is as precise and effective as possible.
You can perform any kind of processing that you require. Use an LLM made available via LivePerson or use your own.
Learn where custom processing of the user’s query fits into the overall search flow.
Use cases
One great use case for this is if you’re supporting cross-lingual queries (for example, the query is in Spanish, but your knowledge base is in English). In this case, you might want to translate the query into the language of the knowledge base that will be searched for answers. Doing so can improve the search results significantly.
Key benefits
- Boost the accuracy and relevancy of answers.
- Support a global reach with multi-lingual capabilities.
Applications using this
Currently, custom processing of the user’s query via Generative AI is only supported by Conversation Assist. Learn how to turn this on in a knowledge base rule within Conversation Assist.
Prompt templates
LivePerson makes available the following related prompt templates in the Prompt Library:
- Query Translation - Conversation Assist (Messaging)
- Query Translation - Conversation Assist (Voice)
Prompt customization
Customizing the prompt? Learn about the variables used in custom processing of user queries.
FAQs
The language of my knowledge base is “Other.” How do I handle this?
When creating a knowledge base, there’s a long list of available languages that you can choose from. But if the content is in a language that isn’t available, you can select “Other” as the language.
In this case, you’ll need to manually edit the prompt for the LLM so that it explicitly specifies the language of the knowledge base. You can’t use the kb_language variable that’s resolved at runtime.