Hi,
I would like to get the contextual prompts from the bot. That means, the bot should intelligently come up with questions based on what the user has entered in the bot earlier. Is it possible for the in-built NLP engine of Kore AI platform to generate prompts based on previously given input from the user?
For example, suppose the user entered that he is facing some error in the website and described the error message. The bot should frame question which includes the relevant error details and ask for where that particular error has occurred instead of a generic question “Where are you facing this error?”.
I have come across this feature called Gen AI Prompt, which will need Open AI license as per my understanding. But I am looking for other alternatives within the platform which do not need Open AI license.
Please share information on this. That would be really helpful.
Thanks & Regards,
Sushma.
Hi @sushma.polakam ,
Thank you for reaching out to Kore.ai Community.
If you do not wish to use Gen AI prompt or LLM, you will have to use script to generate the prompts.
Regarding the example for framing the questions, please refer to the information present in the below documentation and make changes as per your requirement.
Thank you,
Srujan Madderla
Kore.ai Community Team
Hi @srujan.madderla ,
Traits are used to trigger intents/tasks that have just the scripted questions. Can you please clarify on how to leverage traits feature to ask customized questions?
Let me explain the use case that I am working on. When a user gives an input, the bot will be reading the input and ask for missing information contextually. Suppose the system recognized that “Who” is missing, for the input “There is a production issue during the deployment of new code”, then the bot should be designed to ask - “Who is facing this production issue during deployment of new code?”
Please suggest a way to solve this problem.
Thanks.