A
Arleen Dimailig
I'd like to recommend expanding Voice AI call feedback beyond the current thumbs-up/thumbs-down rating system.
1. Five-Star Call Quality Ratings
  • 5 — Excellent
  • 4 — Good, minor improvements needed
  • 3 — Okay, needs improvement
  • 2 — Poor
  • 1 — Critical failure
2. Root-Cause Categories
  • AI Conversation / Reasoning
  • Prompt / Knowledge Base
  • Workflow / Automation
  • CRM Data / Configuration
  • Caller Behavior / External Factors
3. Separate AI Quality From Operational Outcomes
An AI agent can handle a conversation correctly while a downstream workflow fails. These should be evaluated independently rather than assigning the AI an inaccurate negative rating.
4. Enhanced Reporting
Provide dashboards showing quality ratings, recurring problems, root causes, and performance trends by AI agent.
Real-World Example:
Our Voice AI agent correctly recognized a customer's callback request and recorded the appropriate end-call disposition. However, a workflow configuration mismatch prevented the callback automation from triggering.
The AI performed correctly, but the workflow failed. The existing thumbs-up/thumbs-down system doesn't adequately capture this distinction.
A more detailed feedback system would help agencies identify actual problems, improve AI performance, and troubleshoot automations more efficiently.
Thank you for considering this enhancement!
Peoples World