We are developing a customer support Agentic Virtual Agent, and after some tests have found that long conversations cause an unexpected error to occur, after about 7 turns that trigger either a tool call or a knowledge base call a nondescript error message appears and the conversation cannot continue further.
We know that this is not triggered by the genesys rate limit as that error is different, and we know it isn't related to our tool as it is also present in conversations that only have access to a genesys knowledge base.
Is there a different rate or token limit that we are hitting? Is there somewhere to check the logs generated during the conversation?
#AIConfiguration#ConversationalAI(Bots,VirtualAgent,etc.)------------------------------
Diego Prada
------------------------------