Skip to main content
Ciarem tracks your setup for you. AI configuration > Monitoring > Readiness shows a live verdict (“Ready” or “Not ready yet”) with every check that needs attention before your agent goes live, each one linked to the screen that fixes it. The Readiness tab in Monitoring, showing the verdict, the required-to-launch checks, and the recommended checks

Required to launch

These four checks must pass:
  1. Business profile. In Control center > Basic configuration, complete your business context (company, industry, what you offer, who your customers are) using Edit context. The agent uses it to introduce your business correctly.
  2. Active AI agent. At least one agent in your AI agent team must be switched on and ready to answer.
  3. Knowledge base. Add your business information under AI configuration > Knowledge base so answers are grounded in your content, not guesses. See Teaching your agent.
  4. Connected channel. Connect WhatsApp, Instagram, or Messenger so customers can reach the agent. See Connecting WhatsApp.
The Readiness tab also checks the things that make a launch solid rather than merely possible: your agents have skills and knowledge to work with, your knowledge indexed without errors, transfer rules are enabled so conversations can reach a human, someone on your team is available to take handoffs, and your automated tests are passing.

Test it yourself

Before real customers talk to your agent, you should. At the bottom of the Readiness tab:
  • Open agent playground: chat with your configured agent exactly as a customer would.
  • Test knowledge base: ask questions against your knowledge to confirm the answers come out right.
Ask the ten questions your customers ask most, in the language they ask them. Fix what reads wrong by editing the setting behind it, then test again.

Go live

When the required checks pass, the verdict flips to Ready. Make sure the agent toggle in Control center is set to Active, and your agent starts answering on every connected channel.

After launch

From day one, the Live health tab in Monitoring starts answering the second question: is it working correctly? See Is my agent working correctly?, and check the Inbox daily for anything waiting on a human.