In this blog post, I will cover how we built a chatbot (using mock dataset) without prompt but using guardrails. As LLMs are open-ended and can generate responses that may not align with organizational policies, a set of safety/ precautionary measurements are must to maintain trust in Gen AI applications.