AI Configuration Review Checklist

    CategoryAI Employee
    Article typeDocumentation
    Estimated reading time 5 min
    Last reviewedRecently reviewed
    Applies toGEVADE CRM accounts where the feature is enabled
    Access requiredAccess may depend on your user permissions
    Plan/configuration noteFeature availability may depend on your selected plan and account configuration

    Summary

    Use this checklist to verify your knowledge sources, boundaries, handoff rules, and billing settings before deploying AI features.

    Who this article is for

    This article is for administrators conducting a final review before launching AI features.

    Overview

    Deploying AI without proper review can lead to poor customer experiences and unexpected costs. This checklist ensures your AI tools are configured safely and accurately.

    Step 1: 1. Knowledge Source Verification

    Confirm that all connected website URLs, FAQs, and text snippets are accurate, up-to-date, and free of contradictory information.

    Step 2: 2. Tone and Instruction Review

    Check custom instructions to ensure the AI's personality aligns with your brand and that formatting rules (like using bullet points) are clearly defined.

    Step 3: 3. Boundary and Escalation Check

    Verify that restricted topics are explicitly listed, fallback responses are polite, and human handoff triggers are routed to the correct team members.

    Step 4: 4. Channel and Permission Audit

    Confirm which communication channels the AI is active on and ensure that only authorized users have permission to alter AI settings.

    Step 5: 5. Billing and Wallet Check

    Review your prepaid wallet balance and auto-recharge settings to ensure usage-based AI services will not be interrupted.

    Step 6: 6. Simulator Testing

    Run test conversations using the AI simulator. Test basic FAQs, complex questions, and intentionally trigger restricted topics to validate the handoff process.

    Configuration note

    Always test before going live. Contact support if an option is not visible. AI behaviour depends on knowledge sources, instructions, handoff rules, enabled features, connected channels, permissions, configuration, testing, and ongoing monitoring.

    Resolution guidance

    What if the AI fails the simulator test?

    Do not deploy the AI. Review the conversation logs to identify the failure point, adjust your knowledge sources or rules, and re-test.

    Was this article helpful?

    Need account-specific help?

    If your question requires account access, billing review, workflow inspection, message delivery diagnosis, AI response review, or private configuration review, submit a support request with examples and screenshots.