This post investigates the benefits and limitations of prompt engineering in two instances of AI-assisted onboarding relying on large language model (LLM) technology. Of particular interest is how truthful (and therefore reliable) LLM-generated answers turn out to be in the
![]()