Define what the AI should include — and what it must avoid
Answer guidelines allow you to create explicit do’s and don’ts that shape AI behaviour. For example:
- Do: provide clear steps, include timeframes, reference policies
- Don’t: promise refunds, commit to timelines, escalate issues
These rules ensure your automated replies don’t break processes, overcommit, or introduce inconsistencies.
They complement your intents and handling rules, resulting in precise, safe, and professional automation.