1. AI on AI: Who checks your CX?
Your AI does the work. Another AI checks it. So who checks the checker?
- How are we using one AI to check another AI’s work, and what has it caught or missed?
- What does good practice look like: the same AI, a different AI, or a person checking a sample?
- What guidelines and guardrails do we need, so a wrong answer does not reach the customer?